A ids and

R esearch
T ools in
A ncient
N ear
E astern
Studies

JOHN L. HAYES

A MANUAL OF SUMERIAN GRAMMAR

AND TEXTS

e UNDENA

PUBLICA TIONS

Malibu

1990
AIDS AND RESEARCH TOOLS IN ANCIENT NEAR EASTERN STUDIES

Editor: Giorgio Buccellati

This is an introductory pedagogical grammar, designed for readers with

no previous knowledge of Sumerian or its writing system, to be used either with
or without a teacher. It includes a general description of the language and its
writing system, and a series of 22 lessons. Each lesson includes: sign-list and
vocabulary; cuneiform text(s); transliteration, transcription, and translation;
linguistic commentary. The texts used are royal inscriptions of the Ur III period,
presented in photograph or autograph. A certain amount of historical, archaeological,
and cultural background is also included. While primarily meant for students
of Mesopotamia who already are familiar with Akkadian, it is also designed for
students of West-Semitic, who may know no Akkadian. For this latter audience,
emphasis has been placed on transliteration and transcription, to enable the
Manual to be used without learning the cuneiform signs.

Copyright (c) 1990 by Undena Publications

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any
means, electronic or mechanical, i ncluding photocopy, recording, or any information storage and
retrieval system, without permission in writing from the publisher.

Library of Congress Card Catalog Number: 89-51971

ISBN 0-89003-198-3 (cloth) /0-89003-197-5 (paper)

Issued under the auspices of I1MAS-The International Institute for Mesopotamian Area Studies

UNDENA PUBLICATIONS, P.O. BOX 79, MALIBU, CALIFORNIA 90265

 PREFACE

Anyone who has ever tried to learn or to teach Sumerian faces a difficult task. First of
all, knowledge of Sumerian is still at an imperfect stage, with fundamental questions yet to
be resolved. Second, there is a lack of both scholarly and pedagogical tools. Although a
recent descriptive grammar exists, there is no up-to-date sign list or dictionary, and there is
no text-book of any kind. This situation makes it difficult for both student and teacher, and
makes it virtually impossible for someone to learn Sumerian without a teacher.
The aim of this book is to help alleviate this situation. It is a textbook of the Sumerian
language, based on the royal inscriptions of the Ur III period. It is self-contained, so that it
will be of use to students with or without a teacher. It includes a general description of the
Sumerian language and its writing system, and then a series of graduated lessons. Each
lesson contains: sign-list and vocabulary; notes on selected vocabulary; text(s) in cunei
form, either photograph or autograph; transliteration, transcription, and translation; line-by
line commentary on the text. Each lesson concludes with discussions, arranged the
matically, of grammatical issues raised by the text, and of the meaning, function, and
historical context of the text. Later lessons also include supplementary texts for review and
practice, with no new vocabulary or grammar. In each lesson the grammar has generally
been presented inductively from the texts. Finally, there are several appendices, some
treating more general topics, and some serving as reference; the last of these is an index to
grammatical (and other) points.
This book has been designed for a one-semester, three-hour per week class. It can
serve as an introduction to the language for students who will not pursue their study of
Sumerian any further, but it will also prepare students for more advanced work.
Two possible audiences are envisaged. The first is composed of those students who
are comfortable in Akkadian, and who wish to learn Sumerian principally because of their
interest in Mesopotamia. The second is composed of those students who are more
comfortable in West-Semitic, and who wish to learn Sumerian principally because of their
interest in Ebla. The latter audience will either not have studied Akkadian at all, or will
have studied it at some time in the distant past, and may have forgotten much. A certain
amount of material for this latter audience is included which will already be known to those
who are familiar with Akkadian. Throughout, a knowledge of basic linguistic terms and
concepts has been assumed. Since the learning of cuneiform signs often seems like an
onerous chore for those students primarily interested in West-Semitic, the book has been
designed with sufficient emphasis on transliteration and transcription to allow it to be used
without learning the signs.
This book is based on the language of the royal inscriptions of the Ur III period. It is
thus a grammar solely of the written form of the language. It attempts to be purely
synchronic, avoiding a mixture of synchronic and diachronic levels. At the same time,
areas of disagreement about the language are pointed out. Some stress has been placed on
the methodological principles involved in studying a language like Sumerian. Since many
of the problems in understanding Sumerian phonology, morphology, and even syntax are

iii
iv Preface

rooted in difficulties with the script, a certain emphasis has been placed on the nature of the
Sumeriim writing system.
In order to give an idea of the context in which the texts are rooted, some
archaeological, historical, and cultural information is included. Similarly, typological
observations about the Sumerian language have been pointed out, to show that there are
other languages which work in ways similar to Sumerian.
Because of the limited subject-matter of the texts which are used here, not all features
of the language are encountered. Some of these features are touched upon in Lesson 23,
where some alternative views of Sumerian grammar are sketched. Appendix 5 discusses
the ways by which students, including those working alone, can deepen their understanding
of Sumerian. This book will be followed by a second volume, consisting of heavily
annotated extracts from Inanna's Descent. The reading of a major literary text will intro
duce students to a number of problems not encountered in reading the rather stereotyped
texts used in this book.
Appendix 4 is a basic bibliography of the most important and interesting books and
articles on Sumerian. In order for students to become become acquainted with the names of
some of the scholars in the field, a number of modern-day Assyriologists and Sumerolo
gists are quoted throughout the book; all works so quoted are listed in Appendix 4
The genesis of this book goes back to my teaching of Sumerian at the University of
California at Los Angeles. It is a pleasure to thank those who have helped out along the
way. Thorkild Jacobsen was my first teacher of Sumerian; his influence can easily be seen
throughout the book. Sara Denning-Bolle graciously drew the cuneiform signs used in the
sign-lists and those scattered throughout the book; I am especially grateful to her. Barbara
De Marco made a number of useful stylistic observations, and helped in the overall
structure. Several individuals read earlier gestations; I would especially like to thank Daniel
Foxvog, Samuel Greengus, and Stephen Lieberman. Other individuals read certain sec
tions; I thank Denise Schmandt-Besserat and Russell Schuh. James Platt, who studied
from this book, made a number of suggestions. Christopher Walker helped me attain
access to a number of photos from the British Museum. Giorgio Buccellati helped in many
ways, from the initial conception to the final product. And, I would like to thank the staff at
Undena Publications, especially Frank Comparato and Patricia Oliansky. Faults remaining
are my own; I would be very grateful to hear from readers with suggestions for revisions.
I would like to dedicate this book to my mother, for her support and encouragement
over all the years.
 CONTENT S

PREFACE . . . . . . . . . i

INTRODUCTION

Importance of Sumerian . . . . . 1
Difficulties in studying Sumerian . . 1
Historical background and texts used 2

PART ONE: THE SUMERIAN LANGUAGE

CLASSIFICATION . . 5
Linguistic affiliation 5
Dialects . . . . . 5
Typological characteristics 7
Ergativity . . 7
Agglutination . 10

WRITING SYSTEM . 11
External characteristics 11
Original nature . . 12
Internal principles 13
Transliteration 14
Transcription 16

PHONOLOGY 18
Problems . 18
Vowels . . 19
Consonants 20
Other features 23

PART TWO: LESSONS IN SUMERIAN GRAMMAR

Lesson 1 25
Lesson 2 47
Lesson 3 57
Lesson 4 65
Lesson 5 73
v
vi Contents

Lesson 6 79
Lesson 7 85
Lesson 8 95
Lesson 9 101
Lesson 1 0 . 1 09
Lesson 1 1 . 1 17
Lesson 1 2 . 129
Lesson 1 3 . 147
Lesson 1 4 . 157
Lesson 15 . 1 65
Lesson 1 6 . 181
Lesson 17 . 191
Lesson 1 8 . 205
Lesson 19 . 21 1
Lesson 20 . 223
Lesson 2 1 . 23 1
Lesson 22 . 245
Lesson 23 . 257

PART THREE: APPENDICES

Appendix 1 :History . 265

Appendix 2: Mesopotamian sources 273
Appendix 3: Glossary 283
Appendix 4: Bibliography 291
Abbreviations 291
Works cited 291
Concordance of texts 299
Appendix 5: Further work 305
Appendix 6: Topical index 309
 Contents vii

TEXTS and ILLUSTRATIONS

Text 1 : brick of Ur-Nammu - autograph 29

Text 2: brick of Ur-Nammu - photograph 47
Text 2: brick of Ur-Nammu - autograph . 49
Text 3a: brick of Ur-Nammu - photograph 57
Text 3a: brick of Ur-Nammu - autograph 59
Text 3b: brick of Ur-Nammu - autograph 60
Text 3c: brick of Ur-N ammu - autograph . 63
Text 4: votive bowl of Ur-Nammu - autograph 66
Text 4a: brick of Ur-Nammu - autograph 72
Text 5: cone of Ur-Nammu - autograph 75
Text 6: brick of Ur-Nammu - autograph 81
Text 7a: foundation tablet of Ur-Nammu - photograph 86
Text 7b: brick of Ur-Nammu - autograph . 87
Text 7c: brick of Ur-Nammu - autograph 93
Text 8: door socket of Ur-Nammu - autograph 96
Text 8a: brick of Ur-N ammu - autograph 99
Text 9: brick of Ur-Nammu - autograph 1 02
Text 1 0: weight of Shulgi - autograph 1 10
Text lOa: brick of Ur-Nammu - autograph 1 16
Text 1 1 : wig of Shulgi - autograph 120
Text l 1a: brick of Ur-Nammu - photograph 128
Text 12: vase of Ur-Nammu - autograph 131
Text 1 2a: door socket of Shulgi - autograph 144
Text 1 3a: brick of Amar-Sin - autograph . 149
Text 1 3b: brick of Amar-Sin - autograph . 1 50
Text 1 3c: brick of Amar-Sin - autograph . 1 55
Text 14: brick of Amar-Sin - autograph . 1 59
Text 14a: cone of Ur-Nammu - autograph 1 64
Text 15: pedestal of Amar-Sin - autograph 1 67- 1 68
Text 1 6: door socket of Amar-Sin - autograph 1 83
Text 1 6a: bead of Shulgi - autograph 1 89
Text 17: door socket of Amar-Sin - photograph 191
Text 17: door socket of Amar-Sin - autograph 194
Text 18: door socket of Shu-Sin - autograph 206
Text 1 8a: bead of Shulgi - autograph 2 10
Text 19: door socket of Shu-Sin - autograph 2 14
Text 19a: amulet of Amar-Sin - autograph 221
Text 20: brick of Shu-Sin - autograph 226
Text 21a: weight of Shu-Sin - photograph 23 1
Text 2 1b: seal of Ibbi-Sin - autograph . 234
viii Contents

Text 21c: seal of lbbi-Sin - autograph . . . . . . . 241

Text 21d: weight of Shulgi - photograph and autograph 243
Text 22: seal of Shulgi - photograph . . . . . . . 245
Text 22: seal of Shulgi - autograph . . . . . . . 249
Text 22a: seal of Ur-Nammu - photograph and autograph 255

reconstruction of ziggurat ofUr-Nammu 42

remains of ziggurat of Ur-Nammu 42
Stela ofUr-Nammu 43
Old Akkadian brick-stamps 53
figurine of Ur-Nammu . . 92
reconstruction of ziggurat of Nabonidus 107
duck weight . . . . . . . 1 12
door socket of Inanna Temple 145
Neo-BabyIonian pedestal 179
Old Akkadian seal . . . . . 236
 INTRODUCTION

finponBnce of S wnerian

For students of Mesopotamia, the need to study Sumerian is obvious. Alongside

Akkadian, Sumerian is of prime importance for reconstructing many aspects of
Mesopotamian history and culture. However, a knowledge of Sumerian is also useful for
those students primarily interested in Semitic linguistics, and for those interested in biblical
studies.
For Semitists, Sumerian is of importance because of its pervasive influence upon
Akkadian - influence upon the phonology, morphology, syntax, and lexicon. Only through
a knowledge of Sumerian can one differentiate between features of Akkadian which are a
product of its Semitic ancestry, and those which have arisen secondarily under the influence
of Sumerian.
Even though Eblaite has only been known for a sport while, it is clear that its study
will have a profound effect on Semitic linguistics. However, the majority of the texts found
at Ebla are written in pure Sumerian, not in Eblaite. The remaining texts, although written
in the Eblaite language, are couched in a Sumerian writing system which obscures many of
the actual Eblaite forms. This means that a knowledge of Sumerian, especially a thorough
understanding of the principles underlying the Sumerian writing system, is of importance
for research in Eblaite.

Difficulties in studying Swnerian

Sumerian is not as well understood as is Akkadian; a number of features in the

morphology and in the syntax are not clear. Although there has been considerable
linguistic progress in the last two decades, enough still remains unsure so that scholars
often have widely divergent views about Sumerian. Some of the reasons for these
difficulties are summarized here; they will be discussed in more detail in the course of this
book.
(1) Sumerian is not genetically related to any other known language, living or dead.
By contrast, it was discovered early-on that Akkadian was a Semitic language. This genetic
relationship aided early scholars in their reconstruction of Akkadian grammar and
vocabulary. But in the case of Sumerian, there is no such help available.
(2) The writing system of Sumerian only imperfectly mirrors the spoken language; it
does not indicate all the grammatical features which are known to have existed in the
spoken language. This schematic nature of the script makes it very difficult to reconstruct
the morphology.
(3) There are many instances of sentences which seem to differ only slightly in their
morphology or syntax. But with no comparative evidence, and with no native speakers to
turn to, it is difficult to determine what these differences in morphology and syntax may
mean. There are undoubtedly many nuances of meaning which cannot be determined at all.
It has been remarked by Igor Diakonoff, "It is a joke well known among
2 Manual ofSwnerian

Assyriologists that there are as many Sumerian languages as there are Sumerologists"
(1976:99). Similarly, ThorkildJacobsen has recently said:
Knowledge of Sumerian is still in a rudimentary, experimental stage where
scholars differ on essential points, so that translations, even by highly
competent scholars, may diverge so much that one would never guess that
they rendered the same text. ... Scholars have not yet been able to agree on
basic gramma r and its restraints (1987:xv).
In certain ways, however, it is actually easier to study Sumerian than it is to study, for
example, Akkadian. This is because Sumerian does not have (at least, there is not visible) a
great deal of "morphology"; there are not a large number of grammatical forms to learn.
There is nothing like the weak-verb systems of Akkadian and Hebrew, which require a
great deal of sheer memorization. Rather, many students find the difficulties to be more
conceptual in nature: the language works in ways different than English, or other
languages which students are likely to have been exposed to. It is sometimes difficult to
understand some of these principles, and even more difficult to observe these principles in
action.

Historical background and texts used

The texts utilized here are all royal inscriptions of the Ur III Dynasty (approximately
2112-2004 BC), sometimes referred to as the Neo-Sumerian Dynasty. It grew out of the
vacuum left by the collapse of the Dynasty of Akkad, which had been ruled by Akkadian
speaking kings of Semitic stock (approximately 2334-2193 BC).
The Ur III Dynasty was founded by Ur-Nammu, who ruled in the city of Ur from
about 2112 to 2095. He had previously been governor of Ur under the suzerainty of the
king of Uruk, Utu-Hengal; he may have been a relative of the latter. At some point he
declared himself independent. During his rule, and especially during the rule of his son
Shulgi, the territory controlled by Ur expanded, until it reached most of the area previously
controlled by the rulers of Akkad, that is, most of central and southern Mesopotamia. After
three more descendants of Ur-Nammu, the dynasty collapsed in 2004, partially due to
pressures from the intrusion of nomadic, Semitic-speaking tribes. Thus, the Ur III period
lasted a little more than a century; with the fall of Ur, Sumerian civilization, for all intents
and purposes, also fell.
Ur III was a period of relative calm and stability in much of Mesopotamia. Because of
the blooming of Sumerian art and literature, which had been somewhat submerged under
the Semitic dynasty of Akkad, this period is often called the "Sumerian Renaissance".
Towns were fortified, many temples were rebuilt, and canals were dredged; trade with
various foreign countries flourished.
The city of Ur itself, the capital of the Ur III Dynasty, was primarily excavated by Sir
Leonard Woolley, perhaps the most famous of all Near Eastern archaeologists. The
principal results were published by him and others in a series entitled Ur Excavations. Ten
volumes have appeared: Volume I in 1929, and Volume VII in 1976 (Volume X appeared
in 1951). Woolley popularized his results in a one-volume work entitled Ur of the
 Introduction 3

Chaldees (1929). After Woolley's death, P.R.S. Moorey revised and updated the work; it
appeared as Ur 'of the Chaldees' (1982). This is a readable and interesting description of
the city at different historical periods.
Many Ur III texts have been preserved. The vast majority are economic and
administrative; these number in the tens of thousands. Unfortunately, there are very few
texts of what might be called a "historical" nature. There is much that is not known about
such matters as Ur-Nammu's rise to power, the internal politics of the Ur III Dynasty, or
even the physical extent of the Ur III "Empire"; C. J. Gadd refers to the "tantalizing want
of information due to the singular unwillingness of the age to record even the triumphs,
much less the failures, of its kings" (1971:617).
Some original literary texts are also preserved from this period, as well as older works
now committed to writing. Jacobsen says that the kings of Ur Ill, especially Shulgi,
were much concerned to preserve extant older literary works and to
encourage the creation of new ones. The court background of these works is
unmistakable. ... A major portion of Sumerian Literature as we have it traces
back to the court of the kings of the Third Dynasty of Ur, where it was
composed and performed by the royal bards (1987:xii, 277).
The royal inscriptions of the Ur III kings have been the object of study by W. W.
Hallo. According to Hallo's definition, royal inscriptions are texts which "were dedicated
either by, or to, or on behalf of the king" (1962:1). Hallo catalogued these texts, providing
a standard system of reference. He also studied the different sub-types of royal
inscriptions, categorizing them according to their function and according to their form.
These texts range in difficulty, from quite simple to very complex. They also contain a
high degree of formulaity; many of the epithets of the king, for example, occur in a large
number of the inscriptions. Even the phrasing of the verbal expressions is rather fixed.
Since the genre of royal inscriptions existed both before and after the time of Ur III (in
Sumerian and in Akkadian), a knowledge of the Ur III texts gives immediate access to
other similar texts.
There has been much recent discussion about when Sumerian ceased to be a spoken
language. This is not an easy question to answer; there are both historical issues and issues
of general linguistics to resolve. (The subject is further discussed in Appendix 1.) Most
Sumerologists would say that Sumerian was a living spoken language in Sumer during the
Ur III period, although some would say that it was already starting to die out during the
latter part of this period. A minority would say that spoken Sumerian was either pretty far
on its road to extinction, or might even have ceased to be a spoken language by the end of
the Ur III period. Even the proponents of this view, however, would admit that the
language of the Ur III royal inscriptions is "good" Sumerian, unlike some Sumerian of later
periods.
 PART ONE : THESUME IDAN LAN GUA GE

CLA S S IFICA TION

Lin guistic affiliation

S umerian appears to be what is called a language-isolate, that is, it has no genetic

connection with any known language, living or dead. Attempts have been made to link
Sumerian with many different languages - the most popular have been Hungarian, Turkish,
Caucasoid, Dravidian, and the Indus Valley language(s) - but none of these has found
general acceptance. Such attempts have usually been based on surface-level resemblances
with languages which are typologically similar.
A. Leo Oppenheim has pointed out:
The fact that S umerian is a complicated though very well understood
language which cannot be linked to any other known language has created
during the past hundred years a large literature attempting to relate Sumerian
to practically all languages between Polynesia and Africa. The authors of
such studies unfailingly "prove" that either their own language or a language
in which they happen to be interested is related to ancient Sumerian
( 1 97 1 :219).
Sir Gerard Clauson has summed this up: "Sumerian ... has every appearance of being
a 'loner', in spite of numerous attempts to foist relatives upon it, some grotesquely
improbable"( 1 973:38).
The possibility that a connection might be found with some other language is slim.
Any related languages have probably died off without leaving any written records. The
original homeland of the S umerians is unknown, so it is not even clear where its possible
linguistic relatives might be located. Wherever such a homeland might be, it was probably
not in an area where writing developed very early.

Dialects
The S umerians referred to their own language by a term often transliterated as: eme
gir 15. The value of the second sign is not sure, and so the term is variously transliterated as
eme-gi7' eme-lrn, etc., especially in older secondary literature. erne means "tongue" in
Sumerian. The meaning of girl S is unsure. Older scholars thought that it meant "Sumer";
in that case, the term would mean "language of Sumer". More recently it has been argued
that the term means something like "noble, prince"; erne-girl S would then mean "the noble
language". Because of the uncertainties in reading this word, the term "Main Dialect" is
often used instead.
There is also a "dialect" called erne-sal. The meaning of the second element of the
name is uncertain; it may mean "fine, thin". The "status" of this dialect is also uncertain. It
has traditionally been called a "women's language", because it appears in literary texts of the
Old B abylonian period, used by women when speaking to other women. For example, in
the myth "Inanna's Descent to the Netherworld", when Inanna speaks to her aide Nin-

5
6 Manual ofSumerian

Shubur, she does so in Emesal. There is no consistency in this usage; in other texts Inanna
may speak in Main Dialect. Moreover, in texts of the later Old Babylonian period Emesal
is also used for specific genres of text. Certain kinds of lamentations are always written in
Emesal, even though recited by male priests. (Texts in some of these genres were
preserved and even composed in schools for a thousand years after Sumerian had ceased to
be a spoken language.) This use by men makes it difficult to determine exactly what
Emesal is, and whether or not it should be classified as a "dialect".
Emesal is well-attested from the beginning of the Old Babylonian period on.
However, there appear to be at least one or two Emesal forms in the Gudea texts, and there
has been a recent attempt to see Emesal forms in a group of texts written in an unusual
orthography from Tell Abu Salabikh (approximately 2600 BC).
Emesal differs from Main Dialect in phonology and in the lexicon, but not apparently
in morphology. In phonology, the Emesal forms often appear to be older. For example,
the word for "lord" in Main Dialect is len/, in Emesal lumun/. It is difficult to say exactly
what the more original form was; it may have been something like */ewenl or */uwun/.
In any case, the Emesal form appears more conservative than the Main Dialect form.
According to other scholars, however, Emesal forms are linguistically the more innovative;
Emesal forms result from consonants being shifted to a more fronted or to a higher place of
articulation. For example, Main Dialect Igl > Emesal Ib/; Main Dialect Idl > Emesal Iz/,
etc. But there are several exceptions to these general principles, and there are a number of
details of Emesal phonology which are not clear. As an example from the lexicon, the
Main Dialect word for the interrogative "what?" is lana/; the Emesal form is /tal. These
are apparently two etymologically distinct words.
It has been claimed that Emesal shares certain characteristics of "women' s languages"
which occur elsewhere in the world. In particular, women's languages are said to differ
from "standard" dialects in phonology - the women's dialect being more conservative than
the standard dialect - and in the lexicon. More work needs to be done in defining the
characteristics of Emesal, and in comparing Emesal with other women's languages.
Not much is known about geographical variation within Sumerian. The extent of the
Sumerian-speaking area i unsure; Sumerian texts are preserved from only a rather limited
area. Moreover, the nature of the Sumerian writing system makes it difficult to see such
vanatlOn. Only traces can be found, particularly in the later periods. There was
undoubtedly more dialectal variation present than the writing system allows us to see.
Similarly, although Sumerian was spoken over a long period of time, there does not
appear to be much variation before the Old Babylonian period. More differentiation is
noticeable in post-Old B abylonian periods, when Sumerian was no longer a spoken
language. But here the differences may reflect the practices of different scribal schools and
scribal centers, and not differences which were originally in spoken Sumerian.
There are occasional references in late Sumerian texts to what are apparently
specialized languages, or jargons of particular occupations. For example, there are passing
references to eme-utula, "the language of shepherds", and to eme-ma-Iah4-' "the language
of sailors". It is hard to say what these dialects or jargons were like. S imilarly, there are
only passing references to what may be some kind of "literary dialects": erne-gal, "great
 Classification 7

language", eme-sukud, "high language", etc. It is not known what these designations
mean.

T ypological characteristics

There are several ways in which Sumerian works differently than the Semitic or Indo
European languages. Consider the Akkadian sentence, "The king went":

(1) arrum illik

king-NOM VERB

Now, consider the Akkadian sentence, "The king built the house":

(2) Sarrum bItam Ipu

king-NOM house-A CC VERB

In Akkadian, "king" is the subject in both sentences : It is the subject of an intransitive

verb in sentence (1), and the subject of a transitive verb in sentence (2). Therefore, in both
sentences it is put into the nominative case, Sarrum. In sentence (2), "house" is the direct
object of a transitive verb, and so it is put into the accusative case, bItam.
Languages in which the subject of a transitive verb and the subject bf an intransitive
verb are marked one way (called the "nominative" case), and the direct object is marked a
different way (called the "accusative" case), are often called "accusative" languages (or
"nominative-accusative"languages) .
Sumerian, on the other hand, is what is called an "ergative" language. In an ergative
language, what we consider to be the subject of a transitive verb is marked by the "ergative"
case. But, what we consider to be the subject of an intransitive verb, and what we
consider to be the direct object of a transitive verb, are both marked by the "absolute" case.
In some ergative languages the ending for the ergative case, and the ending for the
absolute case, may look completely different. In other ergative languages, the ergative case
will have one marking, but the absolute case will be unmarked. ("Unmarked" can also be
understood as "marked by zero". This can be symbolized by "zero": 0.) In other
languages, there is no case-marking on any of the nouns; rather, ergativity is reflected in the
way that certain elements within the verb cross-reference the case relationships.
In Sumerian, sentences ( 1 ) and (2) would be expressed as follows (Here and
elsewhere, a period is used to separate morphemes; the verb forms have been slightly
simplified) :

(3) luga1.0 i.gin

king-ABS VERB

(4) luga1.e e.0 mu.n.du

king-ERG house-ABS VERB
8 Manual ofSumerian

In (3), the subject of the intransitive verb is marked by .0, the absolute case-marker.
In (4), the subject of the transitive verb is marked by .e, the ergative case-marker, while the
direct object is marked by .0, the absolute case-marker. This fits the definition of an
ergative language: The subject of a transitive verb is marked one way (in Sumerian, by .e),
while the subject of an intransitive verb, and the direct object of a transitive verb, are
marked a different way (in Sumerian, by .0).
Ergativity is a different way of marking the primary participants in a sentence. In an
accusative language, the subject of a transitive verb and the subject of an intransitive verb
fall into one grammatical category; in an ergative language, the subject of an intransitive
verb and the object of a transitive verb fall into one grammatical category. Consider the two
English sentences, "The ball rolled down the hill", and "The boy rolled the ball down the
hill". In English, "ball" in the first sentence is the subject, but in the second sentence it's the
direct object. Yet in each case, it's the ball that is rolling down the hill. In an ergative
language, "ball" would be in the absolute case in both the first and second sentences, and
"boy" would be in the ergative case in the second sentence. In this example, an ergative
language seems to capture our intuitions about the role of the ball in these two sentences
better than does our accusative language.
In the above discussion, the terms "subject" and "object" were used. However, it is
imprecise (and unjustified on theoretical grounds) to use these two terms when talking
about an ergative language. Most linguists prefer to use the term "agent" to refer to the
subject of the transitive verb (marked by the ergative case), and the term "patient" to refer
both to the subject of the intransitive verb, and to the direct object of a transitive verb (both
marked by the absolute case). Thus, in the examples above, "boy" is the agent, and "ball"
is the patient. In practice, it is very difficult to escape using such common terms as
"subject" and "object", especially in unambiguous contexts, even if these terms do not
really fit Sumerian.
There are many ergative languages in the world, belonging to a number of different
language families: many languages in Australia, many American Indian languages, the
Caucasoid languages (for example, Georgian), Basque, to name a few. However, none of
what are sometimes referred to as the "major cultural languages" of Europe are ergative,
and so the concept is unfamiliar.
There are two other important points about ergativity. First, the definition given above
describes what may be called "minimally" ergative languages. However, ergativity can also
be reflected in other parts of a language's grammatical system - it may affect verbal
agreement, cross-referencing of case-markers, coordination and subordination, etc. This
will be discussed in more detail later.
Second, there appear to be very few (if any) "pure" ergative languages. Most (perhaps
all) ergative languages are "split". In certain constructions, the language behaves in an
ergative manner; in other constructions, the language behaves in an accusative manner. In
S umerian, for instance, the perfect aspect functions in an ergative manner, while the
imperfect aspect functions in an accusative manner. That is, Sumerian is split along an
aspectual axis. There are other languages in the world which are split along exactly such an
axis, that is, the perfect aspect functions in an ergative manner, and the imperfect aspect
 Classification 9

functions in an accusative manner. Also, the independent pronouns in Sumerian function

basically on an accusative, not an ergative, basis. Languages of the world show a rather
bewildering variety and complexity in the ways that they are split.
In addition, there are languages which use an ergative - absolute differentiation to
mark semantic distinctions which are not easily made in the Semitic or Indo-European
languages. An oft-cited example is the sentence "We fell" in Bats, a member of the
Caucasoid language family, spoken in Georgia. If the act of falling is purely an accident,
outside of our control, the subject of the sentence is in the absolute case. If we fell as a
result of our own action, the subject is in the ergative case. Other languages use an ergative
- absolute differentiation to mark other kinds of information, such as degrees of animacy.
Because there are very few (if any) pure ergative languages, it is perhaps best not to
think of "ergative - accusative" as a simple binary opposition. C.T. van Aalderen has said
that "One suspects that the whole phenomenon is more a continuum than a set of
oppositions" ( 1 982:27). That is, some languages are closer to one "pole" than to the other.
Several recent linguists, for example, speak of "degrees of ergativity" in different
languages.
In the last twenty years or so, general linguists have shown a great deal of interest in
ergative languages; the bibliography of recent works is vast. In one of the more recent
articles, John Du Bois says:
Seemingly, ergativity stands as a challenge to the view that all languages are
built on one universal archetype. ... Why are there ergative languages in the
world? ... Ergativity ... would seem somewhat perverse in splitting up an
apparently basic category like subject, assigning half its contents to a
contrasting category like object. This perception of unnaturalness is of
course only an index of our failure to apprehend the actual basis of ergativity,
a difficulty which is simply reinforced by traditional grammatical
terminology ( 1 987 :805-7).
It is only somewhat recently that the term ergative has been systematically used for
S umerian. Although some early researchers had intimations that this was how Sumerian
worked (even if all the details were unclear, as they still are), it is only in the last few years
that ergativity has been explicitly discussed in Sumerian. This means that in reading even
fairly recent Sumerological literature, such concepts and terms as "ergative", "agent",
"patient", etc., may not be used at all. The material might be discussed in what would now
be called an ergative model, without use of the term ergative, or in older works the material
might be presented in an accusative model. Moreover, not all scholars believe that
Sumerian functions on an ergative basis. Some S umerologists believe that not enough
evidence has been presented to prove the case, and also believe that there are too many
"exceptions" to the model. Others disagree on the degree to which Sumerian can be said to
be split. Given the complexities of split ergativity in the languages of the world, it may be
that current presentations of ergativity in Sumerian are too simplistic. "Full" proof can only
be forthcoming when there is more secure knowledge of Sumerian verbal morphology.
The first person to apply the term ergative to Sumerian was apparently Viktor
Christian in 1 957, although he used the term a little differently than it is usually understood.
10 ManualofSumerian

Diakonoff ( 1965) sketched the system of ergativity in Sumerian and other Ancient Near
Eastern languages, without explaining the details of morphology. The articles by Daniel
Foxvog ( 1 975) and Piotr Michalowski ( 1980a) viewed Sumerian in an explicitly ergative
framework, while elucidating the verbal morphology. Van Aalderen ( 1 9 82) has explored
some of the theoretical issues in more detail. The grammar by Marie-Louise Thomsen
( 198 4) also follows a split-ergative model. A recent survey of ergativity in Sumerian is by
Gong Yushu ( 1987).

Agglutination

Sumerian is often described as an "agglutinative" language. This term goes back to the
nineteenth century, when linguists attempted to classify the languages of the world into a
few basic types, based solely on typological (not genetic) criteria. For these linguists, the
three most common types of language could be classified as:

Isolating
In isolating languages, virtually every morpheme forms a separate "word". In
Chinese, for example, there are no tense-markers on verbs; such information is conveyed
by separate adverbs. There are also no plural-markers on nouns or verbs; this information
is conveyed by separate number-words.

Fusional
In fusional languages, such as Akkadian or Latin, grammatical morphemes are
expressed through endings on nouns or verbs, and several different morphemes tend to
"fuse" together. Latin amo, for example, means "I love". The /0/ ending on the verb
signals several things: the verb is first person, singular, present tense, indicative mood,
active voice. However, none of the morphemes for person, number, tense, mood, or voice
can be segmented out - they are all fused into the ending /0/.

Agglutinative
In agglutinative languages, as in fusional languages, several grammatical morphemes
are combined into one word. However, the morphemes are distinct from each other; they
do not fuse together. In an agglutinative language, strings of prefixes or suffixes tend to
occur; each affix is formally distinct, and expresses one morpheme. The parade example of
a language of this type is Turkish. In Turkish, the phrase "from his houses" is expressed
as: evlerinden. Ev means "house", ler is the plural marker, in is the possessive pronoun
"his", and den is the postposition expressing the ablative "from". In general, each affix
expresses one morpheme; each morpheme is invariant: ler is the automatic plural marker
for all nouns; den means "from" after any nominal phrase, etc. The morphemes are distinct,
not fused into each other.
Sumerian is similar to Turkish. The verbal phrase, for example, consists of a string of
prefixes, followed by the verbal root, and then a smaller string of suffixes. Each affix
expresses one morpheme, and each affix is (basically) invariant. Nominal phrases can be
very long, with a noun, modifying adjectives and appositives, genitive phrases, etc., with a
 Writing system 11

case-marker at the end of the entire nominal phrase.

The typological scheme presented here has been somewhat simplified. Moreover,
languages only tend to one category or the other; they are not "purely" isolating, fusional,
or agglutinative. English, for example, is largely isolating, but it is also to some degree
fusional. It is occasionally agglutinative in its processes of word formation. In English
words such as "predictability" or "antidisestablishmentarianism", it is fairly easy to separate
several different morphemes, both as prefixes and as suffixes.
Most modern linguists who specialize in linguistic typology are not very interested in
this particular "morphological typology". They believe that such a scheme is not especially
useful, because it does not offer any interesting or helpful intuitions or generalizations
about language. The methodological underpinning of this classification scheme has also
been attacked on several grounds. For example, it was mentioned above that languages do
not usually fall neatly into one of these types. However, since the term agglutinative is still
used in Sumerological literature, especially in popular descriptions of the language, it is
useful to have some idea of what the term means.
The two terms ergative and agglutinative refer to different categories. The ergative -
accusative distinction depends on how the primary participants in a sentence are marked in
relation to each other. The isolating - fusional - agglutinative distinction refers to the
different ways that morphemes are combined into words. In theory, a language can be
either ergative or accusative, and also either isolating or fusional or agglutinative, although
not all of these possible categories seem to occur.

WRITING SYSTEM

External characteristics

In discussing any writing system, there are two factors to consider: the external
characteristics of the script, and the principles behind the script.
Because of the external shape of the signs in the Sumerian script, its writing system is
called "cuneiform". "Cuneus" is the Latin word for "wedge"; the term was coined because
of the most striking characteristic of the script - the fact that the signs are built up of strokes
looking like little wedges. (The term cuneiform was apparently first used by one Thomas
Hyde in 1700. In his Historia religionis veterum Persarum, he refers to "dactuli pyra
midales seu cuneiformei".)
The cuneiform signs were inscribed by means of a stylus probably formed from an
actual reed (such as still grows in modern-day Iraq), by impressing the stylus upon a tablet
of moist clay (or, occasionally, upon other surfaces). The stylus could also be made of
bone, metal, hardwood, or even other material.
The first cuneiform texts discovered were all relatively late, from a period when the
wedge-shaped characteristics of the script were most striking. In the earliest phases of the
script, however, this wedge-shaped character is less pronounced; the script of most of the
Ur III inscriptions in this book does not look nearly as wedge-shaped as do later texts.
12 Manual ofSumerian

The tenn cuneifonn refers solely to the external shape of the individual signs.
Cuneifonn script was adopted and modified by many peoples of the Ancient Near East; it
was used to write Akkadian, U garitic, Hurrian, Persian, etc. However, the fact that these
languages use signs with the same general external characteristics says nothing about their
possible genetic relationship. Sumerian, Akkadian, Hurrian, and Persian, for example,
belong to four entirely unrelated language families. Expressions such as "cuneifonn
language" are occasionally encountered, but this is a rather imprecise way of referring to
one or several languages, which may or may not be related, which use a script with the
same external characteristics.

Original nature
The writing system used for English is an attempt to render speech as closely as
possible. Although English does suffer from numerous archaic spellings, and there are
certain features (such as upper and lower-case letters) which are found only in writing,
writing is basically an attempt to reproduce speech sounds. By contrast, the Sumerian
writing system was never an exact, phonetic representation of speech; it was not "designed"
to reproduce spoken language as such. Rather, to some degree the writing system is only a
mnemonic device, to jog the memory of the writer and reader. The earliest uses of writing
were for administrative texts, which were of a fonnulaic nature, and whose contents were
familiar to the scribes. There was no need to write down what would be obvious to a
scribe who was a native speaker of Sumerian, and who was familiar with the material being
written. When such scribes "read" the texts, they knew how to supply the infonnation not
indicated explicitly in the writing.
Thus, a certain amount of infonnation in the spoken language was not expressed in the
writing. The further back in time one goes, the less the Sumerian writing system expresses
grammatical elements which are assumed to have been present in the spoken language. For
example, the basic graphic shape representing the root for "to build" was originally a
picture of a wooden peg. In the earliest Sumerian, this one sign could be used for any
inflected fonn of the verb: any tense, mood, or person. S imilarly, the expression for "on
that day" in Sumerian was: ud-bi- ("day-that-on"). But in the earliest Sumerian, only the
ud-sign was written; the reader inferred the rest.
As might be imagined, this lack of explicitness in the script can cause much trouble in
interpreting Sumerian texts. Nor is this problem limited to the earliest Sumerian texts; in
late economic texts, for instance, it is often difficult to tell if something is being distributed
"to" or "from" somebody.
As time passed, the scribes wrote more and more down, that is, the writing became
more and more explicit. For example, there is a Sumerian text known as the "Kesh Temple
Hymn", attested in several copies mostly from the Old Babylonian period (dating to around
1 800 BC). In the 1960s, a version of the same text was found at Tell Abu Salabikh, dating
to about perhaps,2500 BC. Unfortunately, only a few lines of the Tell Abu Salabikh
version survive. But if one compares the Old Babylonian version with the Tell Abu
Salabikh version, it can be seen that although the text itself is relatively stable, the Old
Babylonian version indicates more verbal affixes than does the Tell Abu Salabikh version.
 Writing system 13

This increase in explicitness may be connected with the fact that Sumerian was gradually
dying out, and so scribes needed more help in their own understanding of texts.
Thus, a fundamental feature of the Sumerian writing system is its lack of explicitness.
It does not fully represent the spoken language. This has been summarized by Jacobsen:
"The history of Sumerian writing is one of progressively ever greater but never quite
attained adjustment to Sumerian speech" ( 1 957:366 n.1). Similarly, Marvin Powell has
pointed out that "We find traces of its mnemonic character enduring to the very end of the
Sumerian orthographic tradition" (1981 :42 1 ).
A further complicating problem is that the writing system is to some degree
morpheme-bound. There is indirect evidence to show that there were certain phonological
changes which took place in Sumerian, such as contraction, vowel deletion, etc., but these
changes are masked by the script; the script often reproduces the basic morpheme, without
showing the changes which are assumed to have taken place in the spoken language.
The view here presented, that the Sumerian writing system in origin and in practice is
basically mnemonic, has been especially expounded by Diakonoff ( 1976) and Stephen
Lieberman ( 1977).
Internal principles
The script used for writing Sumerian is a combination of "logographic" and "syllabic"
elements. Logographic means that a sign stands for a particular word. For example, the
sign 4 stands for the word utu, "sun"; the sign stands for the word digi!:, "god". The
external shape of many of these signs is clearly pictographic in origin. Thus the sign for
"sun" was originally a picture of the sun rising over a mountain. The sign for "god" was
originally a picture of a star. The original significance of many signs cannot yet be
determined.
The same sign can often have more than one logographic value. Thus, the same sign
can represent diEi!:, "god", or it can represent an, "sky". In general, it is only the context
which determines the meaning of the sign, and its correct reading.
Syllabic signs are used to reproduce a sequence of phonetic elements. For example,
the sign is used to represent the syllable Iga/. This particular syllable can form a
component of several different morphemes: it may be part of the cohortative prefix on
verbs, or part of the ending of a genitive phrase on nouns, etc. The sign in these
contexts does not stand for any particular word; rather, its purpose is to represent the
phonetic sequence I gl -I ai, which may form part of a number of different morphemes.
Syllabic signs can represent several different kinds of segments of consonants and
vowels. Some syllabic signs stand for single vowels, e.g., and i. More common are
signs standing for the sequence consonant-vowel (ba, mu) or vowel-consonant (fill, in).
There are some signs that stand for consonant-vowel-consonant, but these are not common;
instead, the script uses a convention that represents ICVCI by CV-Vc. For example, the
segment Inirl is written by: ni-ir. A writing such as ni-ir does not imply a long vowel;
this is purely an orthographic convention, to reduce the number of CVC-signs which
would otherwise be necessary.
Many signs have more than one syllabic value. Many signs have both logographic
14 Manual ofSumerian

and syllabic values - sometimes more than one of each. The correct value of the sign can
usually only be derived from the context. Signs with more than one value are called
"polyvalent", or are said to have several "readings".
Thus, the Sumerian writing system is both logographic and syllabic. The syllabic
value of most signs derives from a logographic value. For example, the sign in its
meaning as "sky" is pronounced lan/. This phonetic value was then generalized, so that
this sign can stand for the syllable lanl in other contexts.
In general, lexical morphemes are written logographically, and grammatical
morphemes are written syllabically, but this is not always the case. The system is
complicated by the fact that certain syllabic signs tend to be used for certain morphemes.
For example, there is a "conjugation-prefix" on the verb, pronounced Ibi!. There are
several different possible ways that this phonetic sequence could be represented in the
script. In practice, however, the scribes almost always used only one of these possibilities,
the sign . That is, certain morphemes tend to be indicated in only one way, and,
conversely, certain signs tend to be used only for certain morphemes.
In addition to logographic and syllabic signs, there are a few other elements present in
the script. One of these is "determinatives". Determinatives are signs which are used to
indicate the general semantic class to which a following (occasionally a preceding) noun
belongs. For example, almost all divine names are preceded by the sign ; this sign tells
the scribe that "what follows is a divine name". Most names of countries are followed by
the sign -+; this sign tells the scribe that "what precedes is the name of a country".
Determinatives were probably not spoken, even when Sumerian was read out loud. They
were only a feature of the written language.
In other contexts, the cuneiform signs which function as determinatives can also
function as logographic or syllabic elements. For example, the sign can represent digir,
"god"; the sign can represent ki, "country".
To sum up, Sumerian is mostly logographic, and only partially syllabic. Akkadian, on
the other hand, is mostly syllabic, and only partially logographic. Persian cuneiform is
almost entirely syllabic, and Ugaritic cuneiform is basically alphabetic. In practice, people
sometimes confuse the issue, and the term cuneiform is occasionally used to refer in general
to any logographic-syllabic system of writing, but this is wrong; there are many
logographic-syllabic scripts which have existed in the world, which are not cuneiform.
This has been a somewhat simplified discussion of the Sumerian writing system.
There has been much recent discussion about the script, mostly hinging on theoretical
questions, such as the difference between pictographic and logographic, or the degree to
which the script is morpheme-bound.
Transliteration
When citing Sumerian texts, or when discussing Sumerian grammar or vocabulary,
Sumerologists do not generally reproduce the original cuneiform signs. Rather, they cite
the word or passage in transliteration into Latin characters. Transliteration is a sign-by-sign
image of the original written text. It is designed specifically to reflect the actual cuneiform
signs present. By looking at a transliteration, one should be able to determine exactly
 Writingsystem 15

which cuneiform signs occur in the original text (excluding palaeographic niceties).
Transliteration serves several purposes. It is more convenient, quicker, and cheaper to
produce Latin characters than it is to produce cuneiform characters. Also, it provides an
approximate phonetic rendering of the signs occurring in the Sumerian. Since many
Sumerian signs have more than one reading, a scholar, by giving the text in transliteration,
explicitly states his opinion about the reading of a particular cuneiform sign. For example,

the sign can be read Hkur (the name of a god), or im ("wind"), or ni ("self ' ). Based
on his understanding of the text, a scholar decides the correct reading.
There are some complexities of transliteration. It is possible for several different
cuneiform signs to have the same pronunciation. These signs must be differentiated in
transliteration, so that the original cuneiform can be reconstructed from the transliteration.
For example, there are at least four different signs pronounced as /u/. If y were used as
the transliteration for all four signs, it would not be possible to go backward from the
transliteration: Given a transliteration y, one could not tell which of the four possible signs
actually was written in the cuneiform. To obviate this problem, scholars have devised the
following system: The most common (or most important) sign with a particular value is
unmarked. The second most common (or most important) sign with this same value is
marked with an acute accent: g. The third most common (or most important) sign with this
same value is marked with a grave accent: !l. The fourth, and higher, most common signs
with this same value are marked with subscripts: .!4 , liS ' etc. This system is purely
arbitrary; it provides a convenient means to differentiate between signs pronounced alike,
thus enabling us to reconstruct the cuneiform from the transliteration.
This use of the acute and grave accent-marks as "indices" has nothing to do with
pronunciation. They do not indicate anything about accent, nor do they indicate anything
about vocalic length, nor do they indicate anything about tone. They are used instead of a
possible Y2 and Y3 simply because it is easier to type accent marks (at least in Europe) than
it is to turn the typewriter carriage up to make a subscript.
These indices are based largely on frequency. However, these frequencies were
determined on the basis of Akkadian texts, not on the basis of Sumerian texts (for the
simple reason that Akkadian was "discovered" before Sumerian). This produces a certain
inconsistency. In Sumerian, for example, the bi-sign is much more common than the bi
sign. This inconsistency is not really a problem; the only other alternative would have been
to devise a separate system for Sumerian, based on values and frequencies in Sumerian.
But this would have engendered so much confusion and complication that it is far easier to
work with the traditional system.
Confusion arises when indices are used on bisyllabic signs, that is, signs which
represent a segment of two syllables, such as /kala/ or /Urim/. If there is more than one
sign with the same bisyllabic reading, some scholars put the accent-marks on the first
vowel, then continue onto the second syllable if there are several signs with the same
reading. Other scholars, however, begin with the last vowel, moving back to the first.
Either system is prone to mechanical mistakes in printing, and the mere presence of the two
different systems can cause problems in determining what the cuneiform sign actually was.
To mitigate against this difficulty, some Sumerologists do not use acute or grave accent-
16 Manual ofSumerian

marks on bisyllabic signs. Instead, they use a subscript 2 or subscript 3 when necessary.
For example, there are several signs with the value of /kala/. These are differentiated as:
kala, kala 2> kala3' kala4' etc. This is the system followed here. Some recent publications,
including the Pennsylvania Sumerian Dictionary, use subscripts in place of accent-marks,
even on monosyllabic signs. Thus, instead of !!, they use l!2 ; instead of y, they use l!3 '
Determinatives are a feature of the written language, and were probably not spoken.
To indicate that they were not pronounced, they are transliterated with superscript letters:
Xki, tugX, etc. For convenience sake, the determinative for god (the iligir-sign) is trans
literated as a superscript d : dBtar. Because of the typographical difficulties of printing
superscripts, some publications instead print the determinatives on the same print-line,
connected by a period: X.ki; tug.x.
In transli teration, signs comprising one "word" are linked by hyphens: kalam-ma,
illgir-ra-ni, etc. (Determinatives are an exception; no hyphens are used.) As will be seen
below, it is not always easy to determine what constitutes a "word" in Sumerian.
Some Sumerologists use initial capital letters in their transliteration of Sumerian proper
names; other Sumerologists do not. Those who do not use them, consider capital letters to
be a feature particular to the English writing system; since capital letters have no correlate in
the Sumerian writing system, they should not be used in transliteration. Other scholars feel
that since transliteration is an artificial device anyway, there is no harm in using capital
letters, if they help make the text clearer to the reader. This second practice is followed
here.
Finally, it is necessary to say a few words about the typographic conventions used in
transliterating Sumerian. Throughout this book, Sumerian is transliterated by Roman
characters, underlined. The few Akkadian citations used here follow the same system.
However, it is occasionally inconvenient to use the same typographic conventions for two
different languages. To solve this problem, many publications cite Sumerian in Roman
characters, but widely-spaced. Thus, the word for "god" will be transliterated as: digir.
This may seem like a convenient procedure to differentiate citations from the two
languages, but it is prone to produce mechanical errors in printing.
It is frequently the case that it is not known how a particular Sumerian sign (or word)
is to be read. Some scholars elaborate the system just discussed, by presenting such
doubtful or unsure readings in caps. For example, the word for "interest-bearing loan" in
Sumerian is written: 1lUI . It is not sure how the first sign is to be read. For this
reason, the word is often cited as: IjAR-ra. Some scholars do, however, believe that they
now know how to read this word, and so nowadays one is likely to see the reading: urS
ra. That is, wide-spaced Roman is used for the "standard" transliteration of Sumerian, and
caps Roman is used for unsure readings. Not all Sumerologists follow this system,
however, and what is sure for one scholar may be unsure for another scholar.
Transcription
Transliteration is, by definition, a reflection of the written language, and so does not
necessarily reproduce the spoken language well (as we think we understand it). For this
reason, most Sumerologists use some form of transcription in their study of Sumerian.
 Writingsystem 17

Transcription is not used as frequently as is transliteration; it occurs in discussions of

grammar, and appears in scholars' own notes.
Transcription attempts to reproduce Sumerian forms in their approximately correct
phonological and morphological shape, disregarding the omissions, conventions, and
idiosyncrasies of the written language. For example, signs appearing as kalam-ma in
transliteration, will appear as kalama in transcription, since that is probably how the word
was actually pronounced.
There is no "official" or "standard" system of transcription of Sumerian. It tends to be
somewhat personal and idiosyncratic, used by each Sumerologist to enable himself to
understand the language behind the written form. This situation contrasts with that of
Akkadian, for example. In Akkadian there is a standard way of transliterating texts, and
also a reasonably standard way of transcribing them. This can be done for Akkadian,
because scholars are generally confident of their understanding of the rules of Akkadian
phonology and morphology; in general, transcriptions of Akkadian done by different
scholars will be quite similar. In the case of Sumerian, there is much less confidence about
the language. Because the script does not always express all grammatical elements, the
morphology is not always sure. Moreover, there are several different analyses of the
phonetic structure of Sumerian.
The system of transcription used by most Sumerologists is not always transcription in
the precise sense of the term. For example, morpheme boundaries are often indicated.
Also, full forms of morphemes are often indicated, even when it is assumed that some
vocalic or consonantal segment probably dropped. Thus, it is actually a kind of mor
phological transcription.
The system of transcription used in this book is based on the system of Jacobsen, and
is similar to what many Sumerologists use. It is a morphological transcription, in that it
separates morphemes from each other. In this system, morphemes are separated by
periods. Features which are assumed to have been present in the spoken language, but
which do not show up in the written language, are enclosed in parentheses. The different
indices which appear in transliteration are ignored. Thus, will be transcribed as e, and
Urim5 as Urim. Exceptions to this latter rule are sometimes made, particularly for gram
matical morphemes which tend to be written in only one way. Thus, the "terminative"
case-ending is normally transcribed by .e, because it is always written with the e-sign,
and never with the e-sign or the M-sign. Similarly, the "enclitic copula" is normally
transcribed as .am, since it is regularly written by the am-sign, and not by the am-sign or
the am-sign. (Details of these conventions will be discussed below.)
The difference between transliteration and transcription should be kept in mind.
Transliteration is essentially sign-by-sign, with the goal of representing the cuneiform signs
which were used in the original. Transcription is essentially word-by-word, with the goal
of approximating the correct phonological and morphological shape of a word. (In practice,
however, the terms transliteration and transcription are occasionally used promiscuously.)
Transcription is important, because transliteration alone masks too many
morphological and phonological issues. Only a consistent transcription can reveal a
thorough understanding of the language of the texts. Some of the simplest inscriptions, for
18 Manual ofSumerian

example, could be translated without knowing much Sumerian, simply from a knowledge
of Akkadian and of simple vocabulary; a transcription reflects the structure of the language
hidden beneath the written form.
At certain times in this book, the purely phonemic structure of Sumerian will be
stressed, ignoring any morphological considerations. In that case, normal linguistic practice
will be followed, and the item will be put between slashes, e.g., Ikalama/.
Thus, our understanding of Sumerian may be reflected in three different ways: a
transliteration, reflecting the written shape; a phonemic transcription, reflecting the
pronunciation; and a morphological transcription, reflecting our understanding of the
pronunciation and morphology.

PHONOLO GY

Problems

It is not easy to reconstruct the phonological system of Sumerian, or the precise

pronunciation of any of its sounds. There are two main reasons for this problem. Since
Sumerian is a language-isolate, there is no comparative evidence to provide help.
Moreover, most of the evidence for Sumerian phonology has been filtered through the
Akkadian phonological system; Sumerian phonology is seen through Akkadian eyes. For
instance, it is quite likely that the word for "son" in Sumerian was pronounced Idomu/,
with an initial lol-quality vowel. But Akkadian does not have an lol-quality vowel, and
hence no 101 -sign, and so this word is spelled out in syllabic Akkadian as: du-mu. If there
were only Akkadian evidence, it might never even be known that Sumerian had an 101-
quality vowel. Thus, the picture of Sumerian of the Ur III period (21 12-2004 BC) is
actually based on Akkadian of the Old Babylonian period ( 1894- 1595 BC), and later.
(Similarly, much knowledge of Sumerian grammar derives from the interpretations given to
it by Akkadian-speaking scribes and scholars; this topic is discussed in Appendix 2.)
Likewise, very little is known about the historical development of Sumerian
phonology. Sumerian was spoken over a period of several centuries (and was used as a
written language for even more centuries). The phonological system of Sumerian at the
time of, say, Tell Abu Salabikh and that of the time of Ur III may have been significantly
different.
To some degree, more is known about the value and pronunciation of Sumerian
grammatical morphemes, than about Sumerian lexical morphemes. This is because
grammatical morphemes are mostly written syllabically, while lexical morphemes are
usually written logographically. Without the evidence of lexical lists (Appendix 2), it is
quite difficult to fix the value of a logogram. For the same reason, it is occasionally
possible to see phonetic change through the course of Sumerian in grammatical
morphemes, but it is more difficult to see such changes in lexical morphemes.
The upshot of this is that Sumerian probably possessed sounds which Akkadian did
not, and which can only be determined using a variety of indirect evidence. Because of the
difficulty of dealing with this indirect evidence, there have been several different
 Phonology 19

reconstructions of the Sumerian phonological system. These reconstructions differ both in

the number of phonemes present in Sumerian, and in the value attributed to certain
phonemes.
In practice, however, most Sumerologists do not try to exactly reproduce the sounds
of Sumerian. Rather, they use the standard values known from Akkadian. Thus, virtually
all transliterations of Sumerian will use the value dumu for "son", even though this is one
of the clearest cases where an 101 -quality value can be postulated for Sumerian. Similarly,
it is sure that Sumerian had a velar I lJ I , which did not exist in Akkadian. The sign i ,
for example, represents I lJ ul , the velar nasal followed by an lul -quality vowel; this is the
morpheme for the first person singular possessive-suffix on nouns. But the normal value
of this sign in Akkadian is Imu/. Therefore, many Sumerologists transliterate this sign as
mu, e.g., lugal-mu, "my king". Other scholars, however, transliterate this sign as g!! I O> e.g.,
luggl-g!! IC! Still others, who wish to be more precise, in fact transliterate this sign as lJ!!lO '
or as some typographical equivalent, such as 1O ' g!!lO' etc.; for example, luggl-g!! l C! This
means that transliterations of Sumerian will differ somewhat from scholar to scholar. The
transliteration used here will reflect the conventional method of transliteration used by most
Sumerologists, even if this reconstruction is somewhat shaky and incomplete.

Vowels
Sumerian had at least the following vowels:

i u
e
a

The precise phonetic value of these vowels, particularly the I el, is unsure.
Many scholars also believe that Sumerian had an lol-quality vowel, but since no 101
existed in Akkadian (at least on the phonemic level), there is only indirect evidence to
reconstruct it. It is very difficult to determine whether any particular Sumerian word had an
101 -quality vowel or an lul -quality vowel; its existence has been established for only a few
cases. Under the assumption of the existence of this 101 -quality vowel, the vocalic system
of Sumerian is more symmetrical:

i u
e 0
a

Other Sumerologists have posited other vowels, such as both an open lel and a closed
le/. Others have posited the existence of nasalized vowels, but the exact number and
quality of these varies from one scholar to another: /i/; le/; /i I and la/; /i I, lal and lel,
etc. Claude Boisson ( 1988) has investigated various reconstructions of the phonemic
system of Sumerian, in comparison with what is known about language in general. He
feels that if Sumerian possessed only four vowels, then the vowel normally represented as
20 Manual ofSumerian

lel was more likely IEI than le/. He also feels that none of the systems of nasals which
have been posited for Sumerian is likely.
It is not sure if there was a phonemic distinction between short and long vowels; this
cannot be told from the script. It has been postulated that there were no originally long
vowels in Sumerian, but that they did arise through vocalic contraction, in particular the
contraction of final root-vowels with initial vowels of suffixes.
As discussed above, in practical terms most transliterations of Sumerian usually only
reflect the vowels known from Akkadian; that is, the four vowels listed above.

Consonants
Most analyses of Sumerian would include the following consonants:

b p m
d t n
g k IJ
z s
b
1 r

(For ease in printing, the consonant indicated above as b is often simply transliterated
as h, without the "dish". Since Sumerian does not have a "simple" IhI, there is no
ambiguity in this usage.)
Virtually all Sumerologists accept the existence of the velar nasal IIJI (although some
scholars prefer to speak of a palatal nasal, and others have seen more complex phonemes,
such as IIJm/). When Sumerian words containing this phoneme are loaned into Akkadian,
it is usually (although not always) reflected as !!g. For example, , "kind of priest"
(Lesson 2 1 ) appears in Akkadian as ang'y.
Transliterations of this phoneme vary. In older works, and in many contemporary
works, it may simply appear as g. Some recent works use g, or some typographical
equivalent (g, etc.). It will be transliterated here as g, in cases where it is assumed by most
Sumerologists to be present. With many words, however, it is not known whether a
phoneme is IIJ /, Igl, or even 1nl or Iml, and so some variation in the transliteration of
certain words appears. For example, the verb "to go" is understood by some
Sumerologists to be Iginl, but by others to be Iginl (or Igen/).
Many Sumerologists believe that Sumerian had a phoneme usually symbolized by
Idr/; its exact phonetic significance is unsure. Its existence has been proven in only a few
cases. Because of the difficulties of proving its existence in specific words, it is usually not
indicated in transcription; instead, in the standard sign-lists and in most transcriptions it is
reflected as g.
Several other consonants have been posited for Sumerian: Ihl, Iwl, Iy/; two (or
more) types of /11; two (or more) types of Ir/; a labiovelar Ikw/; a pre-nasalized labial stop
I mb/; etc. Since none of these sounds exists in Akkadian, the evidence for their existence
in Sumerian is indirect at best, and individual Sumerologists have their own preferences.
 Phonology 21

Transliterations of Sumerian do not normally try to reproduce these disputed phonemes.

As a typical example of a reconstruction of Sumerian phonology, it may be instructive
to present that postulated by Lieberman:

e i b p m
a 0 d t n z
u g k g
z s
b
1 r f

In the tables above, certain consonants are indicated as differing only in voice: Ib/
Ip/; Idl - It/; etc. It is not in fact sure what differentiated such pairs; Lieberman explicitly
says that the distinction he marks as Ibl - Ipl was not one of voice. Some Sumerologists
have speculated that the difference was one of aspiration; this is not an uncommon view
today. Boisson, for example, says: "A correlation of aspiration seems to be the only
hypothesis with a high probability of success" ( 1988:25). Other Sumerologists have
speculated that the difference was one of glottalization.
There does not appear to have been a phonemic distinction between short and long
consonants; it is not in fact sure if long consonants occurred at all.
One of the thorniest questions in Sumerian involves the status of word-final and
syllable-final consonants. According to most Sumerologists, certain consonants, when in
word-final position, were not pronounced. For example, the root for "dais" is Ibarag/,
with a word-final Ig/. However, unless this Igl was followed by a vowel, it was not
pronounced: this word would have been pronounced as Ibara/.
The word-final consonant in a root is usually referred to by the German term
"Auslaut". Thus, it is said that the word for "dais" (pronounced Ibarag/) had a "g
Auslaut", or the word for "to live" (pronounced Itil/) had a "l-Auslaut".
The consonants which were regularly not pronounced in word-final position are called
"amissable" consonants. Those which were pronounced in word-final position are called
"non-amissable". (These terms are apparently peculiar to Sumerologists; they are not used
by general linguists.)
Sumerologists differ among themselves about which consonants were not
pronounced. Some believe this affected all consonants, although perhaps not "to the same
degree". Others believe that it affected a smaller number of consonants (although no two
lists of such consonants seem to agree exactly). Also, it is not known if the amissable
consonants were not pronounced in word-final position only; most Sumerologists believe
that they were not pronounced in any syllable-final position. Arno Poebel, for example (the
real father of Sumerian grammar), states that "As a rule, an amissable consonant is dropped
whenever it stands at the end of a word or syllable" ( 1935: 147). Similarly, Samuel Noah
Kramer says: "All final consonants in Sumerian are amissable. ... The term 'final
consonant' as here used includes the consonant at the end of a syllable as well as the one at
the end of a word" ( 1 936: 1 9).
22 ManualofSumerian

The existence of amissable consonants is certainly not impossible. There is a close

parallel in French: In spoken French, word-final consonants are not pronounced (under
certain conditions), although they still appear in the written form. A few Sumerologists,
however, are not convinced of the existence of amissable consonants. They interpret the
problem as being orthographic in nature.
The reason this question is still unresolved is because of ambiguities in the writing
system. At various points in this book, different pieces of evidence will be cited, some of
which seem to indicate that word-final consonants were pronounced, and some of which
seem to indicate that word-final consonants were not pronounced.
The existence of amissable consonants means that the cuneiform signs which represent
words with these amissable Auslauts have two values: a "long" value, which includes the
amissable Auslaut (e.g., kalag, UrimS' !ill , and a "short" value, which does not (kala, UriS '
ill . With some signs, the long value and the short value have different indices, e.g., tU
[with diacritic] and !i [without diacritic]. This annoying situation is partially due to the fact
that indices were originally assigned on the basis of frequency in Akkadian, not Sumerian.
Some scholars transliterate Sumerian using basically only the long values; others
transliterate Sumerian using basically only the short values. Other scholars use both, the
choice being determined by syllabic conditions: the short form if word-final (or syllable
final), the long form if not. Others are less consistent, using a mixture of long and short
values. This latter practice is particularly true of less recent Sumerological literature, where
one finds a mixture of transliteration principles, based primarily on customary readings of
the cuneiform signs. Such customary readings have arisen from the piece-meal growth in
understanding of Akkadian and Sumerian. For example, in 1940 Kramer published an
edition of the "Lamentation over the Destruction of Ur". This is a Sumerian poem, some
436 lines long, bemoaning the destruction of Ur at the end of the Ur III period; it was
written probably about a century after its destruction. In his Introduction, Kramer says that
"The time is not yet ripe for a thorough and scientific overhauling of the Sumerian system
of transliteration". Therefore, he "deems it best to follow the more or less established
usage". In this system,
In the case of signs representing roots that end in a consonant and may have
either the long or the short value (e.g., the signs for ill!( Q), "to call", du(g),
"good", etc., which may be read either pad, dug, etc. or !ill, dul O ' etc.) the
transliteration uses the longer value in spite of the fact that the shorter is
scientifically the more correct. Only in cases such as !!(Q), "day", and a(g),
"heart", where the shorter value has become more or less standard, is that
value used in our transliteration, although the inconsistency in transliterating
the signs for ill!(g) and du(g) as pad and dug while giving those for !!(g) and
a(g) as 14 and M is only too patent (1940:6).
Kramer is obviously irked by this inconsistency, but feels that there is nothing he can
do about it. Although he wrote this passage almost fourty-five years ago, some editors of
Sumerian texts still follow such customary usage. A compromise made by some
Sumerologists is to put the Auslaut within parentheses, e.g., kala(g). However, if the short
and long forms have different indices, this can create confusion; some scholars transliterate
 Phonology 23

as ti(1), others as ti(1). In this book, all word-final consonants have been consistently
transliterated (and transcribed) .

Other features
There were undoubtedly other features in the spoken language, which the writing
system only hints at. There is only marginal evidence, for example, to determine word
stress, and it will not be dealt with here. Similarly, there is only the most indirect evidence
for sentence-intonation.
Because of what is claimed to be a large number of homonyms in Sumerian, it has
several times been argued that Sumerian possessed phonemic tones. Diakonoff, for
example, says: "Sumerian was certainly a tonal language, or else the many homonyms
would have made spoken Sumerian quite unintelligible" ( 1983:86). However, the evidence
is indirect and slight. In fact. many words which earlier Sumerologists believed to be
homonyms have been shown to contain different Auslauts, and so are not actually
homonyms.
 PART lWO: LESSONS IN SUMERIAN GRAMMAR

Lesson 1

This first text is a royal inscription of Ur-Nammu, the founder of the Ur III Dynasty
(ruled 2 1 12-2095 BC).

Sign-list and vocabulary

In this and subsequent sign-lists, the signs are loosely organized according to function.
Determinatives are first, followed by proper names, nouns, verbs, and syllabic signs.

Determinative preceding divine names (DNs). Transliterated by a superscript "d":

W kiDeterminative following geographical names (ONs). Transliterated by a superscript

"ki":

}>- Nanna Nanna (DN, masc)

Nammu Nammu (DN, fern)

Ur_dNammu [PN], masc)

JId r Ur-Nammu (personal name

4 It=tJ Urims (Uris ) Ur (ON)

r nin lady, mistress; "lord"

an heaven

}>-/Tl>-J ur man, warrior

fg{rit> lugal king

M house

p dti to build

4 na

Tt !

P ni

25
26 Lesson 1

}-J!-T ma

Jm ke4

.( mu
.(

Especially for those who are primarily interested in West-Semitic, it is not always easy
to master cuneiform signs. In certain ways, however, it is easier to learn the signs of this
period than the signs of later periods. In later periods, the repertoire of possible sign
shapes becomes quite reduced, so that (superficially) the signs of the Neo-Assyrian period,
for example, all look very similar. In the earlier periods, however, the signs are much more
distinctive, making them easier to learn.
However, one problem in studying the signs of the early periods is the occasional
wide variation in external shape of the signs. For example, the sign for , "house", looks
rather different in Text 2 than it does in Text 1 . This variation is due to several factors:
nature of the writing surface, different scribal traditions at different scribal centers,
individual idiosyncrasies of handwriting, etc. The sign-lists and vocabularies attempt to
produce the basic or essential shape of each sign; the signs in the autographs are
reproduced exactly as published.

Notes
The Notes discuss some of the more important vocabulary items. Often, reference is
made to Akkadian words which were borrowed from these Sumerian words. This practice
is open to methodological criticism, since Akkadian is not Sumerian, and there is no reason
to assume that Sumerian words always kept exactly the same meaning when placed into an
Akkadian context. But since normally much more is known about the Akkadian term than
about the Sumerian term, it is still useful to examine the Akkadian equivalents.

Nanna The city-god of Ur. The large temple-complex at Ur discussed below was sacred
to him in particular. He was associated with the moon; nanna in fact means "moon".
In Akkadian, the word nannaru occurs, glossed by the CAD as: "luminary, light (as
poetic term, an epithet of the moon god and !Star)". This Akkadian word may be some
kind of blend or contamination between the Sumerian word nanna and the Akkadian root
nawanI.
Because of this Akkadian word, some earlier Sumerologists believed that the
Sumerian word had an Ir/-Auslaut, and so the name sometimes appears as Nannar.
However, there seems to be no inner-Sumerian evidence which would indicate such an
Auslaut.
The moon-god was also referred to as Zuen; this problem will be further discussed in
Lesson 1 3.
The Mesopotamian scribes interpreted the cuneiform sign expressing his name as
J:
consisting of two signs: the SeS-sign ( or }> ) followed b the ki-sign ).
Therefore, in older works the name is sometimes transliterated as: SeS-ki. More likely,
 Lesson 1 27

however, the second element was originally the na-sign, functioning as a phonetic
complement of some kind.

Nammu Not much is know about this goddess. However, she is described as "the mo
ther who gave birth to heaven and earth", and as "the primeval mother, who gave birth to all
the gods". It is thus possible that at one time she played a more important role in Sumerian
cosmogony.
The cuneiform sign which represents this name can also be read engur, which lexical
texts equate with the Akkadian apsu, the "watery deep" (see Lesson 14). The cuneiform
sign may be an abstract representation of this deep.
In some older Sumerological works, the two readings of this sign (Nammu and engyr)
were not clearly differentiated. Therefore, the name of the founder of the Ur III Dynasty
sometimes appears as Ur-Engur, or Ur-Gur.

UrimS In English, "Ur". One of the more famous cities in southern Mesopotamia; the city
after which the Ur III period is named. The name of the modern site is al-Muqayyar.
The etymology of the name Urims is unknown. It is also not known how these two
particular cuneiform signs (presumably, the e-sign followed by the ab-sign) came to
represent the name.
Urims is the long value of the sign. The short value is variously transliterated as Uri ,
Uri2, Uri3 , or Uris . The oscillation in diacritics illustrates the problem of diacritical marks
on bisyllabic signs.
The sign-lists in this book give the long value first, followed by the short value.
Because both are encountered in Sumerological literature, it is necessary to know both
values, even though this seems like a totally unnecessary burden upon the student.
Sometimes, the name is written e-unug and not e-ab, in which case it should pro
perly be transliterated as Urim2.
The English equivalent, "Ur", derives from the Old Testament f'ur kasdlmf, "Ur of
the Chaldeans". Exactly how the Hebrew f'urf derives from the Sumerian fUrimf is
unsure.

nin In general, the Sumerian word for "lord" is en; the feminine equivalent, "lady", is nin.
(It is not impossible that the two words are etymologically related.) However, in older
Sumerian nin can also be used to refer to masculine entities. Perhaps at one time the term
was genderless. In the Ur III period, this usage can be considered an archaism.

ur The usual interpretation of this word is something like "man; warrior, hero". In
bilingual lexical texts, ur is glossed as amelu, "man", and as kalbu, "dog". ur with the
meaning "dog" is not uncommon in Sumerian texts. However, ur meaning "man" seems to
occur only in personal names; it does not have this meaning in actual texts (although the
compound ur-sag, "hero", presumably "man-head", is common).
One might guess that the ur-sign was originally a picture of a dog or some kind of
beast, but even the earliest attestations of the ur-sign do not look very animal-like.

l ugal Etymologically, a compound of lu "man" and gal "great". This word is further
28 Lesson 1

discussed in Lesson 7.

According to 1.1. Gelb, "The Sumerian word has several meanings: a) a dwelling
house, even a room b) palace, temple c) family, clan d) household. The same meanings
occur also for the Akkadian bItum" ( 1979b:2). In the sense of "temple", it can refer either
to one particular building, or to an entire temple complex consisting of several buildings.
In very recent secondary literature, it is occasionally transliterated as: 'a.

dd Although du occasionally means "to build" de novo, it more often means "to rebuild".
It is especially frequent when describing the rebuilding of temples which had fallen into
disrepair. Usually, it is difficult to tell in any particular text whether du means "to build" or
"to rebuild"; this can only be resolved by historical or archaeological data.
Gelb adds that "It is clear that when a ruler writes of having built a temple for a certain
divinity, he means not only that he erected a temple, but also that he provided it with all the
necessary means of social and economic support" ( 1 979b:3).
 Lesson 1 29

Text 1

~
trr ~
m w P>
*" re:
5

. r=Cl
rmEf

---
W P --

t>
30 Lesson 1

Notes: autographs

When obtainable, photographs of the texts used in the Lessons have been included.
This has not always been possible or desirable, and so most of the texts are presented as
"autographs". In Assyriological parlance, autograph refers to the hand-copy done by a
modern Assyriologist, to imitate the cuneiform. The quality of autographs can range from
very accurate to very poor. To quote Lieberman,
It is, of course, patent that the "autographs" of all copyists are not equally
reliable. Their objectives, ranging from an exact reproduction including
every scratch on the tablet to a highly abstract conventional representation of
the original (some Assyriologists are even known to have produced "copies"
from their transliterated notes) as well as their individual skills and abilities
make the value of their copies diverge (1977: 67).
It is only through long experience that one gets a feel for how accurate certain
Assyriologists are (or aren't) in their autographs.

-Writing practices

Both Sumerian and Akadian are written from left to right across the writing surface.
(The earliest Sumerian texts were inscribed in vertical columns, read from right to left.)
Most royal inscriptions are subdivided into "lines", marked by an actual line drawn or
impressed on the writing surface. The use of such lines in Sumerian (and Akkadian) is to
some extent dependent on the genre of text; royal inscriptions, for example, use them
regularly. Many literary texts use them, but just as many do not.
There is some oscillation in the use of the word line. This particular text was divided
by its scribe into seven units, but the fifth of these units actually contains two rows of text.
In order to be precise, some Sumerologists use the term "case" or "register" to describe the
units physically demarcated by the scribe, and the term "line" to describe the actual rows of
signs. Thus, in this text case 5 has two lines. Although this is a very handy distinction,
most scholars, will, in fact, simply use the term line to mean both line or case, especially in
k
unambiguous contexts; this is the rocedure followed here.
In line 5, the determinative 1 begins the second line within the case. There are six
cuneiform signs in this particular expression. It would have been physically impossible to
put all these six signs on one line, so the scribe put them on two lines. If he had put the ki
sign with the Urims -sign, there would have been too much empty space on the second line
of the case. By indenting the second line of the case, the signs representing the ON are
grouped in close proximity to each other.

Transliteration Transcription Translation

1: dNanna Nanna For Nanna,
2: nin-an-na nin.an.a(k) the "lord" of heaven,
3: nin--ni nin.ani.(r) his "lord" -
4: Ur_dNammu Umammu Ur-Nammu,
 Lesson 1 31

5: lugal-UrimSki-ma-ke4 lugal.Urim.ak.e the king of Ur -

6: -fl-ni e.ani.0 his temple
7: mu-na-du mu.na.(n.)du.0 he built.
Commentary
1 . Nanna i s the name of a god; Nammu i s the name of a goddess. Sumerian has no gender
system; there are no special markers for either inherently masculine or inherently feminine
nouns. In most cases, one word may apply to either gender. For example, dim may mean
either "god" or "goddess". In other cases, the masculine and feminine seem to be formed
from different roots. In a few other cases, Sumerian adds the word for "female" (munus)
after a noun. For example, dumu can either mean "son" (masculine) or "child" (masculine
or feminine); dumu-munus is specifically "female child", hence "daughter".
2. nin is used here to refer to the male god Nanna. For convenience sake, nin in such
contexts may be translated as "lord".
S umerian has no definite or indefinite article. For example, can mean "a house" or
"the house".
nin.an.a(k) forms a "genitive phrase". The formation of the genitive in S umerian is
quite different from the formations in Semitic or in Indo-European. In Sumerian, in a
genitive phrase consisting of two nouns, the "possessor" follows the "possessed". The
two nouns themselves are not formally marked, but the second noun is followed by the
"genitive marker" .ak. For example, "the house of the king" is: e.1ugal.ak; "lady of
heaven" is: nin.an.ak. (Genitive phrases of more than two nouns will be discussed later.)
The form of the genitive marker is lakl following a consonant (in transcription, .ak)
and Ikl following a vowel (in transcription, .k).
Ikl is one of the amissable consonants discussed under Phonology. As such, when in
word-final position, it does not show up in the writing system. As stated above, most
Sumerologists believe that the reason such consonants do not appear in writing, is because
they were not pronounced. A minority of scholars, however, believe that they were
pronounced, and their absence is purely an orthographic problem. In the morphological
transcription used here, the Ikl is transcribed within parentheses: .a(k). This transcription
shows that the Ikl does not appear in the script.
This genitive phrase is written nin-an-na, which is interpreted as: nin.an.a(k). One
might have expected a writing of the type *nin-an-fl. However, Sumerian generally avoids
writing word-final (and to some degree, syllable-final) single vowels. Instead, the writing
system prefers to graphically reduplicate the consonant immediately preceding the word
final vowel. Thus, in this case, Sumerian writes the na-sign - graphically reduplicating the
preceding 1nl.
The principle of graphically reduplicating a preceding consonant is common
throughout all periods of Sumerian. It is purely a property of the orthography; it does not
mean that S umerian pronounced a double consonant here.
To summarize, nin-an-na represents the genitive phrase: nin.an.a(k). A genitive
phrase of two nouns is formed by adding the genitive marker after the second noun. The
32 Lesson 1

genitive marker is lalc! after a consonant, Ikl after a vowel. Ikl is one of the amissable
consonants, and hence does not appear in writing in word-final position. The lal of the
genitive marker is usually contained within a sign which reduplicates the consonant
immediatelypreceding the/a/.
3. nin-i!-ni = nin.ani.(r). S umerian has a set of suffixes to indicate pronominal possession.
They are referred to as "possessive-suffixes" or "pronominal suffixes". .ani is the
possessive-suffix marking third person singular. The fonns of the first and second per
sons, and of all the plurals, are discussed later.
S ince Sumerian has no gender system, .ani can mean either "his" or "her". However,
Sumerian does have remnants of what is usually referred to as a distinction in "animacy".
Human beings are "animate"; things and animals are "inanimate". In the case of the
possessive-suffix, .ani is only used to refer to animate antecedents; an entirely different
fonn (.bi) is used to refer to inanimate antecedents (corresponding to English "its").
After a consonant, the suffix appears as: .ani. After a vowel, it appears both as: .ani,
and as: .ni. For example, "his house" can appear as both -i!-ni and -ni; in the Ur III royal
inscriptions, the fuller spelling is much more common .
.ra is the case-marker for the dative case. Its fonn is Iral following a consonant and
Irl following a vowel.
Case-endings in Sumerian work differently than they do in the Semitic or the Indo
European languages. In Sumerian, case-endings occur at the end of an entire nominal
phrase. A nominal phrase can vary in size. Minimally, it can consist of a single noun. It
can also consist of a noun with a possessive-suffix, or with an adjective, or with an
embedded genitive phrase, or even with a long series of appositives. In this particular case,
the nominal phrase spans lines 1 to 3. It consists of: a divine name (Nanna); an appositive,
consisting of a genitive phrase (nin.an.a(k ; a second appositive, consisting of a noun with
a possessive-suffix (nin.ani). The dative case-marker .r comes at the end of this entire
phrase. This can be diagrammed as: [Nanna nin.an.a(k) nin.ani].r. This is, in general, the
way all case-markers work in S umerian (and, even more generally, in agglutinative
languages).
The dative case is primarily used in S umerian to express an indirect object; for
example, "He gave it to the king". It is also frequently used (as it is here) to express a
benefactive, that is, the person on whose behalf an action was perfonned. In such cases, it
can be translated by "for".
The case-marker Irl is not written here. Its presence in spoken S umerian is shown by
the fact that it is actually written in other (mostly later) inscriptions. In these other
inscriptions, there are fonns such as: digir-ra-ni-ir, "for his god" = digir.ani.r (following the
nonnal convention that CV-VC stands for ICVC/, that is, ni-ir = Inir/). In the body of
texts in this book, .r first appears in Text 14, an inscription of Amar-Sin, the grandson of
Ur-Nammu.
It is not known why the Irl is not written; this is discussed in Lesson 14. The
situation is different from that of the genitive marker. The Ikl of the genitive marker is an
amissable consonant, and so is regularly not written. But Irl is apparently a non-amissable
consonant, and does occasionally appear in the writing.
 Lesson 1 33

4. ur followed by the name of a deity is a very common way to form personal names in
Sumerian, in all periods of the language. In Text 1 , Ur_dNammu occurs; in Text 1 9a, Ur
dLamar. Such names are genitive phrases, meaning "man of DN" or "warrior of DN".
The name then is to be understood as: Ur.Nammu.(k), with the genitive marker taking the
form Ikl after a preceding vowel. However, there is some disagreement about the presence
of the genitive marker in proper names. Some Sumerologists believe that in proper names
the genitive marker was deleted. Thus, this particular name may have been pronounced as
lumammu/, and not as lumammuk/. Other Sumerologists however, do not believe this to
be so. The first practice has been followed in this book, and thus this name has been
transcribed as Umammu, not as Ur.nammu.(k). In translation, the most common
Assyriological practice is to give the name as Ur-Nammu.
5. l.Y.gal-Urims ki-ma-ke4 = lugal.Urim.ak.e, "king of Ur". Because the genitive marker
follows a consonant (here, Im/), its full form (/ak/, with initial la/) is used.
When the genitive marker is directly followed by a vowel, the Ikl is pronounced, and
shows up in the writing (recall that in such phrases as nin.an.a(k), the Ikl is word-final,
and hence does not show up in the writing).
The .e is the marker of the ergative case, as discussed under Ergativity. As do all
case-markers, it comes at the end of the entire nominal phrase. The nominal phrase here
consists of a personal name, Ur-Nammu (a genitive phrase in origin), and an appositive
consisting of a genitive phrase, lugal.Urim.ak. This may be diagrammed as: [Umammu
lugal.Urim.ak].e.
The ergative case-marker .e marks what we would call the active subject of a transitive
verb, or, in more appropriate terminology, the agent. (Because of inconsistencies in
terminology, however, this .e is sometimes referred to as "agent", "agentive marker" or
"ending", "subject", "transitive subject", "ergative marker", etc.)
The cuneiform signs do not reflect well the morphology of Sumerian here. In
transliteration, the signs are: lugal-Urimski-ma-ke4' In morphological transcription, this is:
lugal.Urim. ak.e. The rna-sign reduplicates the final Iml of Urims, and includes the lal of
the genitive marker. The ke4 -sign includes the Ikl of the genitive marker, and the I el of
the ergative case-marker. Thus, both the rna-sign and the ke4 -sign represent segments of
two different morphemes. This use of the ke4 -sign is very frequent; it is the sign normally
used for the combination of segments of the genitive marker and the ergative case-marker.
Not much is known about the syllabic structure of spoken Sumerian, but it may have
been closer to the written form than to the morphological transcription. This line may have
been syllabified something like: Ilu-ga-Iu-ri-ma-ke/. If so, the written form is actually
closer to the presumed syllabic structure of Sumerian than it is to the morphemic structure
of Sumerian.
The use of hyphens in transliteration varies to some degree from scholar to scholar.
All S umerologists would use hyphens in the wo.rd Urims ki-ma-ke4' Some would put a
l
hyphen between lugal and Urims : lugill-Urim l-ma-ke4' In this latter case, hyphens are
being used to link all the signs which form the entire nominal phrase. Others use hyphens
only between signs belonging to one word. It is not always easy, however, to define
"word" in Sumerian.
34 Lesson 1

6. --ni = e.ani.0, "his temple". As in line 3, .ani is the third person animate possessive
suffix. The antecedent is ambiguous; it could refer to Ur-Nammu, or it could refer to
Nanna. From other texts it is clear that .ani refers back to Nanna.
The .0 is the case-marker for the absolute case. This case indicates what we would
call the direct object of a transitive verb, or, more appropriately, the patient. There is,
however, not a great deal of consistency in nomenclature, and so such terms as
"accusative", "direct object marker", etc., are commonly used.
The nominal phrase here is quite short, consisting of the noun , and the possessive
suffix .ani: [e.ani] .0.
7. mu-na-du = mu.na.(n.)du.0, "he built". This line contains the verbal phrase. The verb
in Sumerian works rather differently than the verb in the Semitic or Indo-European
languages. A finite verb form in Sumerian consists of a series of verbal prefixes, followed
by a verbal root, then followed by a smaller series of verbal suffixes. Certain of these
affixes are obligatory, while others are optional. Because of the general uncertainty of
Sumerian grammar, the precise number of prefixes occurring before the verbal root is
unsure. The view presented here might be called "minimalist". Alternative interpretations
will be discussed later.
The entire sequence of verbal prefixes occurring before the verbal root is usually
referred to as the "verbal chain". The first prefix to appear in this chain is an optional
"modal-prefix" (also referred to as a "mood-marker"). Modal-prefixes are used for such
sentence types as cohortative, jussive, subjunctive, etc. A "normal" declarative sentence is
in the indicative mood, which is unmarked. The verb in line 7 is indicative, and so there is
no modal-prefix.
The second position is occupied by the "conjugation-prefix". There are some half
dozen conjugation-prefixes. These prefixes are among the most mysterious features of
Sumerian; it is not known exactly what information these prefixes convey. This means
that it is not known, for example, what the difference in meaning is between a finite verbal
form with the conjugation-prefix mu and one with the conjugation-prefix i. Such variation
occurs in the texts, but it is not known what this variation implies.
Needless to say, there are several theories about the function of the conjugation
prefixes. They may be connected with time: indicating whether events are near or far
(temporally, or even emotionally) relative to the speaker. They may have to do with space:
indicating whether events are near or far (spatially, or even emotionally) relative to the
speaker. At times, they seem to correspond to a polite - familiar distinction.
It is probable that the conjugation-prefixes convey nuances which are not normally
conveyed in English. This means that even if it were understood what the conjugation
prefixes meant, it would not be possible to translate them readily into English, except by an
elaborate periphrasis. (Jacobsen, for example, believes that the conjugation-prefix mu is
used "To indicate 'closeness' to the speaker if by closeness we understand not only
closeness in space and time but also emotional closeness, empathy, involvement"
[ 1 965: 4 37].)
In practice, Sumerologists ignore the conjugation-prefixes; they are not reflected in
translation. Writing in 1972, Maurice Lambert said: "Today, the prefix does not exist for
 Lesson 1 35

the translator of S umerian, it is only an object of study for the grammarian" ( 1 972-3:97).
The problem of the conjugation-prefixes cannot be solved here. In subsequent texts,
the various conjugation-prefixes will be pointed out, and the possible kinds of information
which they may be conveying will be discussed.
Text 1 uses the conjugation-prefix mu. This conjugation-prefix is very common in the
Ur III royal inscriptions. In fact, almost all past-tense verbs in main sentences in the Ur ITI
royal inscriptions use the conjugation-prefix mu.
The next set of prefixes are the (mostly) obligatory "dimensional-prefixes". There is
nothing comparable to these forms in Semitic or Indo-European. They "cross-reference"
(or "resume" or "register") the case relationships appearing in the various nominal phrases
in the sentence, with the exception of the agent and patient. In the verb in line 7, the
dimensional-prefix .na cross-references the dative case marked by .r in line 3.
Most earlier studies of S umerian stated that the dimensional-prefixes were obligatory,
and that there was a one-to-one relationship between case relationships and dimensional
prefixes: every case relationship is resumed by its dimensional-prefix, and conversely
every occurrence of a dimensional-prefix implies a corresponding case relationship
somewhere in the sentence. While this one-to-one correspondence may have been valid for
"pre-historic" Sumerian, in actual historic Sumerian the situation is not so neat. Gene
Gragg has made a detailed study of the dimensional-prefixes in the Old Babylonian literary
texts; he states that they "function independently of concord to a much greater extent than
has been recognized by current theories" ( 1973a: 10).
The dimensional-prefixes often seem unnecessary or redundant, because they do not
convey any new information; rather, they "merely" cross-reference the already-present case
relationships. However, all languages have a certain amount of built-in redundancy, to help
cope with the possiblities of information being garbled or lost. Many other languages
cross-reference case relationships, in various ways.
The nominal phrase in the dative is the only nominal phrase (except those indicating
the agent and the patient) in the sentence, so only one dimensional-prefix occurs. If other
nominal phrases were present, they would also be resumed. Thus, it is possible for there to
be one, two, or three dimensional-prefixes in one verbal chain; that is, the dimensional
prefixes are cumulative. (The longest attested sequence appears to be four dimensional
prefixes in one verbal chain.) There is a hierarchical order to these prefixes; the dative, for
instance, always comes first. Not all such rules, however, are understood; in addition, there
are certain morphophonemic changes which are not clear. These complications will be
discussed later.
Following the dimensional-prefixes comes a (probably) obligatory prefix, the
"personal-affix" (there is no generally-accepted term). These forms have been much
discussed. They apparently cross-reference the agent and the patient, although this is not
completely certain.
In the case of a verb in the past tense, the personal-affix in this position cross
references the agent. Thus, in Text 1 , the personal-affix .(n) cross-references the agent
marked by the ergative case-marker of line 5.
The form of the third-person singular animate personal-affix is: .n. As will be seen
36 Lesson 1

later, the personal-affix has different forms for first and second person, and also different
forms for inanimate agents.
This particular prefix always occupies the position closest to the verbal root.
However, this prefix frequently does not show up in the writing. The reason for its
absence is not as clear as that of, for example, the dative case-marker. The dative case
marker is not normally written in texts from the early stages of the Ur III dynasty (nor in
earlier texts), but it begins to show up frequently in texts from the time of Amar-Sin on.
Thus, scholars are reasonably confident that the Irl of the dative case-marker is present,
even if not written; its later appearance is the result of a change in orthographic practice.
The rules governing the presence and absence of the personal-affix 1nl are, however, not
so clear-cut; it is not often written even in later texts.
The presence or absence of .n cannot simply be correlated with a dimension of time.
In the Gudea texts, for example, forms both with and without .n occur, with no obvious
rules governing their distribution. And in later S umerian, forms also occur both with and
without the .n. This means that rules cannot yet be determined for the presence or absence
of 1nl in the script, and it is not in fact sure at what level such rules would apply. The rules
may be purely orthographic; there seem to be other cases in Sumerian where syllable-final
nasals are not expressed in writing. Or, the rules may be phonological; the 1nl may have
dropped early, leaving a nasalized vowel, which could not adequately be represented in the
script. More probably, there may be a complex set of morphological and syntactical rules
governing deletion of 1nl; it has been posited, for example, that 1nl is only used (and so
only expressed in writing) to resolve possibly ambiguous cases.
Partially for convenience sake, I have assumed that the personal-affix .n is always
present, unless there is a specific reason for its absence. Hence, it is transcribed as: (n.).
This presumed consistency must be taken with a grain of salt.
After all these obligatory and optional prefixes, comes the verbal root, du in this
particular case. The root in Sumerian appears to be invariable. There is nothing like the
complicated inflection of Semitic or Indo-European roots for person and number (the only
inflection for person is in the personal-affix position, immediately before the verbal root; a
limited inflection for number occurs in a set of personal-affixes after the verbal root).
There is no canonical shape of the root. Roots of the syllabic shape ev and eve are
perhaps the most common, but roots of other syllabic structures are frequent.
After the verbal root, there occur a number of optional affixes, not all of which are
well-understood. Some of these affixes are used to express modal and other nuances, such
as potentiality, irrealis, etc.
For a verb in the past tense, the most important affix which occurs in this position is
the personal-affix which cross-references the patient. The personal-affix which cross
references a third-person singular patient can be represented by zero, .0. Thus, the patient
in this sentence (e.ani.0) is resumed by a .0 after the verbal root. This means that the
patient is marked by .0, and that it is cross-referenced by .0. This may vaguely seem like
cheating ("nothing resumed by nothing"), but there are theoretical justifications for this
interpretation.
Thus, the agent and the patient are resumed differently: The agent is resumed in the
 Lesson 1 37

position immediately before the verbal root, and the patient is resumed in the position
immediately after the verbal root. The term personal-affix is used to refer to both affixes.
To sum up, the verbal phrase in Sumerian normally consists of: an optional modal
prefix (the indicative is unmarked); an obligatory conjugation-prefix, whose function is
unclear; one or more basically obligatory dimensional-prefixes, which cross-reference all
case relationships (except that of the agent and patient); an obligatory personal-affix, which
in the past tense cross-references the agent; the verbal root; an obligatory personal-affix,
which in the past tense cross-references the patient; other optional affixes.
This particular verbal form may be summarized as follows:

mu na . (n.) du . 0
(1) (2) (3) (4) (5)

(1) conjugation-prefix
(2) dimensional-prefix cross-referencing the dative
(3) personal-affix cross-referencing the agent
( 4) verbal root
(5) personal-affix cross-referencing the patient.

The verb in line 7 was translated as past tense, without any discussion. Sumerian has
two sets of verbal forms. The difference in function between the two is somewhat unsure.
Some Sumerologists believe that the difference was one of tense (past - present-future);
others believe that it was a difference of aspect (perfect - imperfect); and others believe that
it was a difference of Aktionsart (punctual - durative, etc.). For convenience sake, they will
be referred to here as aspects.
Akkadian scribes gave names to these two aspects. One aspect they called bamm
("quick"), and the other they called maru ("fat"). There is some evidence that the S umerian
word for bamm was Y4 ' and the word for marii was niga; the original meaning of these
two words is not sure. The terms bamn! and marii are frequently used by modem Assy
riologists when referring to these verbal forms in Sumerian.
In the Ur III royal inscriptions, it does seem that basically the bamt.!! is used for
actions which occurred in the past, and the marii is used for actions in the present and
future. That is, the two seem more tense-like than aspect-like. But this may be due to the
relative simplicity of these inscriptions.
The bamn! form is unmarked; it is the citation form (the form given, for example, in
the Vocabularies). As will be seen later, the marii is formed from the bamt!! in several
different ways, and the systems for cross-referencing the agent and patient in the bamt!! and
in the marii are quite different.

Discussion: structure

Having examined this inscription with a fine-tooth comb, let us now consider the
structure of the inscription as a whole. If all appositional noun phrases are grouped with
their head nouns, and their functions are labeled, we see:
38 Lesson 1

[Nanna, nin.an.a(k), nin.ani] .(r) benefactive

[Umammu, lugal. Urim.ak].e agent
[e.ani] .0 patient
mu.na.(n.)du.0 verb

The dative marked in .r is resumed by the dimensional-prefix .na; the ergative marked
in .e is resumed by the personal-affix .n; the absolute marked in .0 is resumed by the
personal-affix .0. This is a rather aesthetically satisfying system; as will be seen later,
however, things often do not hang together so neatly.
Second, let us look at the word order:

BENEFACTIVE - AGENT - PATIENT - VERB

(1) (2) (3) ( 4)

This particular order is actually somewhat different from standard Sumerian syntax.
In more standard Sumerian, the word order is:

AGENT - PATIENT - COMPLEMENTS - VERB

(1) (2) (3) (4)

or: AGENT - COMPLEMENTS - PATIENT - VERB

(1) (2) (3) ( 4)

The difference in word order between standard Sumerian prose and that of the royal
inscriptions is in the position of the benefactive. In royal inscriptions, the benefactive is
almost always fronted; this gives added emphasis to the deity on whose behalf some act is
being commemorated. In English, the difference might be reflected as "For Nanna, Ur
Nammu built his temple", instead of "Ur-Nammu built his temple for Nanna".
Hallo's investigation of the structure of the Ur III royal inscriptions showed that their
style is very formulaic. A typical inscription is composed of the following elements, almost
al ways in the same order:

(1) A benefactive phrase, giving the name of the deity, with optional epithets;
(2) An agentive phrase, giving the name of the builder or donor, with optional
epithets;
(3) A patient phrase, describing the object built or donated;
(4) A verbal phrase, highly stylized and formulaic.

-Terminology

As does any discipline, Sumerology has engendered its own host of technical terms,
such as Auslaut, amissable, etc. Some of these terms are peculiar to Sumerologists; they
are not standard terms familiar to general linguists. Unfortunately, some of these terms are
used in ways which cause general linguists to take umbrage.
The term verbal chain is used here to refer to the series of prefixes which occur before
the verbal root. Other people use the term to include the entire verb: prefixes-root-suffixes.
 Lesson 1 39

Similarly, the tenn nominal chain is sometimes used to refer to a nominal phrase.
Sometimes both the verbal chain and the nominal chain are subsumed under the category
"Kettenbildung".
The term conjugation-prefix, in particular, is misleading, because these elements have
nothing to do with conjugation, as this tenn is usually understood. However, this is the
only term used by Sumerologists.
There is no standard tenn to refer to what is called here the dimensional-prefix; the
most common tenn is probably dimensional infix. This use of the tenn infix, however, is
often irksome to general linguists, who use the tenn to refer explicitly to an affix placed
within another morpheme; an example would be the It/ in the Akkadian Bt stem, or the
Arabic Eighth Form.

-ke4

The value of theiMJ sign as ke4 was deduced by Kramer in 1936. There is no native
grammatical tradition which gives this value; the Akkadian lexical lists give the values of
this sign as g and k4 (in addition to such values as kid, etc.) Kramer reasoned that the
only way to make the Sumerian writing be consistent with our understanding of the
morphology of the Sumerian genitive was to posit a reading ke4 ' even if the lexical lists do
not give this value. Virtually all modern scholars have accepted his reasoning. (However,
even this seemingly well-established fact of Sumerian grammar has recently been
questioned, by Lieberman. He believes that the genitive marker was /ag/, not /ak/, but he
has not yet published his reasons for doubting the conventional interpretation.)

-Animacy

As was mentioned when discussing the possessive-suffix .ani, Sumerian has traces of
an animate - inanimate distinction. This distinction is also seen in the personal-affix of the
third-person bamt!!-transitive verb, where .n marks an animate agent, but .b marks an
inanimate agent (rather a rare occurrence). This animate - inanimate distinction does not
carry through all aspects of the grammar.
The tenns animate and inanimate are those traditionally used by linguists, even if this
means that animals are called inanimate (Jacobsen prefers the tenns "personal" and "non
personal"). In fables, however, animals are usually treated grammatically as animate.

-Conjugation-prefixes

Lambert was quoted above, to the effect that the conjugation-prefixes are simply not
tran slated. This is because it is not known what infonnation they convey, and the odds are
that their function has no easy equivalent in English. Edmond Sollberger has said:
Their true rOle is so distinctively Sumerian, they express ideas so alien to our
languages, that not only is there no consensus on the nature of their function,
but we simply ignore them without impairing, or so it seems to us, our
understanding of the text. There is no other translation for mu-gar and i-gar
than "(he) placed", although it must be pretty obvious that had there been no
difference there wouldn't have been two prefixes . ... It is legitimate to posit
40 Lesson 1

that a certain verbal fonn implies that the action is perfonned by the subject
wishing to indicate that his goal, though within his immediate perception,
remains without his actual sphere of physical contact; it is another thing to try
and express that in one good English (or even Gennan) word (1973:160-61).
F.R. Kraus has criticized this view of Sollberger: "Sollberger's opinion, that Sumerian
texts can be understood without paying attention to the verbal prefix, is valid for a certain
kind of text, but is certainly not valid for legal documents" (1958:83 n.47).

- Conjugation

The fonns of the bamm-transitive verb in the singular are listed here. This and other
paradigms should be understood as reflecting Ur III morphology, in Ur ITI orthography.
The model verb used is sar, "to write", with the conjugation-prefix mu.

first person singular mu-sar mu. 0.sar

second mu-sar mU.e.sar
third animate mu-sar mU.n.sar
inanimate mu-sar mu.b.sar

The fonn of the first person is somewhat unsure. The fonn of the second person is
more sure, because the .e sometimes shows up in the script. Similarly, the fonns of the
third person are "sure", because of the occasional presence of .n and .b in the script.
In this section, the personal-affixes .n and .b have been discussed as markers for the
third person. Earlier, it was said that they cross-reference the agent. S trictly speaking, they
cross-reference a third person agent. A first person agent ("I") is cross-referenced by . 0,
and a second person agent ("you") is cross-referenced by .e. In other words, one can
understand the personal-affixes as cross-referencing the agent, or as marking the person of
the verb; in Sumerian, these are two different ways of describing the same thing.

- bamm and marii

The first person to recognize that bamm and marii were used as native grammatical
tenns was Heinrich Zimmern, in 1 885, although he did not know what they meant. Paul
Haupt was apparently the first to give these words their etymologies as "quick" and "fat",
in 1932.

- Typology

The Introduction discussed morphological typology, which is one attempt to broadly

categorize the languages of the world into a limited number of types. That particular
scheme of classification is somewhat out of favor, partially on theoretical grounds, and
partially oil the grounds that it does not offer more enlightening insights about language. A
more revealing scheme of linguistic typology is called "word order typology" (although
"constituent order typology" might be a more apt tenn). This scheme examines the "basic"
order of the major constituents in a sentence. In English, for example, the most typical
order is: subject-verb-object. Hence, English is said to be a S- V-0 language. Sumerian,
 Lesson 1 41

on the other hand, i s a S-O- V language.

S-O- V languages pattern alike in several ways, not just in word order. For example,
very few S-O- V languages have prepositions. Instead, they use case-endings at the end of
nominal phrases, that is, "postpositions".
Also, in most S-O-V languages, adjectives follow their head noun, not precede. As
will be seen in Lesson 2, this is also how Sumerian works. Thus, in many ways (not in
all), S umerian is a typical S-O-V language.
G. Haayer ( 1986) discusses some of the characteristics of S umerian in light of the
universal tendencies of language. He points out, for example, that "Most ergative
languages have SOY basic word order", and "The combination of ergativity and
postpositions in a single language points almost invariably to S OY basic word order"
(1986:80).

- Function of text

Let us now look at the function and Sitz im Leben of this particular text. Hallo has
divided the Ur III royal inscriptions into five categories, based on typological criteria:
standard, building, votive, weight, and seal inscriptions. Text 1 is a building inscription;
examples will occur of all the other four types. Building inscriptions are defined by Hallo
as "monuments that became integral parts, whether functional or decorative, of the
buildings which they commemorated" (1962:8).
The building inscriptions are further subdivided on the basis of the type of object they
were inscribed on: bricks (the most numerous of all royal inscriptions), foundation
deposits, door sockets, and clay cones. Examples will be seen of each. Text 1 was
inscribed on a brick, forming an actual part of the masonry of a building. Building
inscriptions in general were not designed to be read by the builder's contemporaries; rather,
they were designed to be read by future rebuilders of the building, most likely future kings.
Ultimately, these buildings and their accompanying inscriptions can be thought of as
attempts by rulers to attain some form of immortality. (Text 16 is a door socket with two
inscriptions. One is of an early ruler of Ur [about 2400 BC]; the other is of a ruler of the
Ur ill period. The door socket was evidently uncovered during rebuilding carried out in
the Ur III period, and was re-used.) As will be discussed in Lesson 2, often several copies
of the same inscription are found.

- History

Throughout Mesopotamian history, temples were built, repaired, modified, or virtually

entirely rebuilt. During the Ur III period, there were many specifically royal building
projects. Building inscriptions of Ur-Nammu have been found at Ur, Eridu, Larsa, Nippur,
and Uruk. He was responsible for building (and rebuilding) the large sacred area at Ur,
consisting of several structures. The most famous of these is its ziggurat, the best
preserved ziggurat in all of Mesopotamia: its base measures some 60 x 40 meters. It was
repaired by several later Mesopotamian rulers. (In Lesson 9, Woolley's reconstruction of
Nabonidus' rebuilding of this same ziggurat is pictured.)
The following drawing is Woolley's reconstruction of the ziggurat. The following
42 Lesson 1

photo is of the remains now standing; the condition of these remains is partially a result of
modem reconstruction of the site.

THE Z ICCU RAT OF UR,.-NAMMU RESTORED .

. "
uf ':.:
" '
" --
----
-- --
-- '--
----
--'--
--'

. . .
 Lesson 1 43

While the sacred area as a whole was dedicated to Nanna, he also had his own coun in
front of the ziggurat, and other buildings sacred to him. The entire sacred complex was
known as the E-ki-nu-Ml; the ziggurat was known as the E-temen-ni-guru3 (see Lesson
9). Both tenns are of uncertain etymology. The brick containing Text 1 fonned pan of a
temple known as the -bur-sag ("mountain temple").
One of the more famous pieces of Ancient Near Eastern art is known as the "Stela of
Ur-Nammu". It was found in a very fragmentary state in Ur, scattered throughout the
Nanna temple complex; it may have been destroyed during the Elamite sack of Ur in 2004
BC. It depicts a number of symbolic activities, mostly obscure to us, but apparently shows
Ur-Nammu himself carrying building tools (his name appears on a floating fragment of the
stela). This stela has been known since the 1920s, but restoration work is still on-going.
(A very interesting discussion is in Canby 1987.)
44 Lesson 1

In Lesson 7, a photograph is reproduced of a figure which represents Ur-Nammu

himself (somewhat stylized) in his role as builder.
Discussing the function of the Mesopotamian ruler in this role as builder, Wolfgang
Heimpel says:
The ruler in Mesopotamia, when building for the gods, manufactured the
first brick himself, sprinkled the foundations with precious materials, laid the
foundation box, mixed some of the mortar, and led the celebrations of
dedication. The best sources for these ceremonies are the building
inscriptions of Assyrian and Neo-Babylonian kings and the cylinders of
Gudea. The latter contain the most detailed information which is couched in
poetic language and presents us with many difficulties of interpretation
(1987:205).

-Literature

In addition to his well-documented role as builder, Ur-Nammu has become famous as

the promulgator of the world's first attested law code, the "Code of Ur-Nammu". At least
three copies of parts of the text are known, but all are heavily damaged. The largest
fragment was found at Nippur. In 1 98 1 a fragment of the Code found at Sippar was
published. Basing himself on this new fragment, Kramer suggested that the "author" of the
Code was not Ur-Nammu, but rather his son Shulgi; this view has won general acceptance.
Ur-Nammu was also the subject of several literary works. These include "The
Coronation of Ur-Nammu", a kind of self-laudatory hymn, and "The Death of Ur-Nammu
and his Descent to the Netherworld", in which his premature death on the battlefield is
lamented.

- Proper names

Most recently, Miguel Civil ( 1985:27) transliterates the name of the founder of the Ur
III Dynasty as Ur-Namma, instead of the usual Ur-Nammu. He bases himself on
attestations of the name in syllabic orthography. He suggests that the original form of the
name was a theoretical IUr-Namnam/. Jacobsen also now reads the original form of the
divine name as Namma, but derives INammal from Inin inim/, "lady female genital";
INammu/ is a later form ( 1987: 155 n.5).
As will be discussed in Appendix 2, a number of bilingual lexical lists have been
found at Ebla. Names of gods occur several times in these lists. For Nammu, the Eblaite
equivalent is given as: i-nu bf!.-mi-um. This is somewhat difficult to understand. F.M.
Fales thinks that the Eblaite expression might mean "venemous tooth" ( 1 984: 176). It is
hard to square such a description or epithet with what is known about Nammu.

-Titulature

Many of the appositive phrases describing the king in these inscriptions are actually
titles, occurring in many inscriptions. (Although sometimes it is not possible to tell if an
adjectival phrase is a title or not.) Much work has been done in determining the origin of
certain titles, their relationship to parallel Akkadian titles, their falling into desuetude, etc.
 Lesson 1 45

The principal work on this topic is by Hallo: Early Mesopotamian Royal Titles: A Philo
mic and Historical Analysis (1957). This was followed by M.-J. Seux in 1 967, who
studied in particular the individual words occuring in Sumerian and Akkadian titles.
The title used in Text 1 , "King of Ur", was used by all five kings of the Ur III
Dynasty.
 Lesson 2

Text 2 is a second building inscription of Ur-Namrnu. It was inscribed on a brick,

fonning part of the Inanna temple (see Lesson 9).

Sign-list and vocabulary

AT Inanna (DN, fern)

Il.. .JJ AJJ)J Ki-m-gi Sumer (ON)

i93: Ki-uri Akkad (ON)

l:.1 nitab (nita) man, male

47
48 Lesson 2

kalag (kala) to be mighty

ID
g

Notes
Inanna The daughter of Nanna. She was the Sumerian goddess of love and fertility, of
the morning and evening star, and to some degree of war; she had other sides as well. She
may have absorbed some of the attributes of originally independent deities. Later equated
with the Akkadian Ishtar, in some ways she was the most important goddess in the
Mesopotamian pantheon. Because of her rather fiery temperament, and the manifold
aspects of her personality, she is perhaps the most interesting of all Mesopotamian deities.
She was worshipped in many cities, but especially in Uruk, where she was the tutelary
goddess. Her principal temple at Uruk was the Eanna -an-na = e.an.a(k), "house of the
sky/heaven").
The reading of her name is much disputed. It is also transliterated as: Inana, Innin,
and Ninni6' The original pictographic meaning of the cuneiform sign is also uncertain. Her
name is usually interpreted as: nin.an.a(k), "Lady of the sky/heaven". This is also how the
Akkadian scribes understood her name. Jacobsen believes that Inanna was originally the
"numen of the communal storehouse for dates". He thinks that the / ani-component of her
name meant "date-clusters": "Her name ... would appear to have meant originally 'the lady
of the date-clusters'" ( 1957 : 1 08); later, her name was "re-interpreted" as "lady of the
sky/heaven".

Ki-en-gi This ON is always written syllabically. The etymology is unsure; this is dis
cussed below. The word ended in a /r/, not reflected in the script. The Akkadian equi
valent of Kiengi was Sumeru. This Akkadian word may be a dialectal pronunciation of the
word Kiengi(r). The English word "Sumer" is usually thought to derive from the Ak
kadian form.
The first appearance of Ki-en-gi is in an inscription of Enshakushanna of Uruk (who
ruled approximately 2432-2403 BC), who refers to himself as: en-Ki-en-gi l!!gal-kalam
ma, "the lord of Sumer, the king of the land".

Ki-uri The etymology is unknown. It is not impossible that the ki-element was originally
a determinative.

nitab The basic meaning appears to be "male"; it can often be loosely translated as "man".
The Akkadian equivalent is zikaru, glossed by the CAD as: " 1 . male (human and animal),
2. man, 3. ram".

kalag The Akkadian equivalent verb, dananu, is translated by the CAD as: "to become
strong". The verbal adjective, dannu, is translated as: " 1 ) solid, strong, hard, heavy, thick,
massive, fortified, steady, loud, 2) legitimate, binding, reliable, 3) strong, powerful, mighty,
great, 4) fierce, savage, difficult, dangerous, serious, grave, obstinate, bad, tyrannical,
harsh, pressing, urgent, essential, imperative".
 Lesson 2

Text 2

* tH bJHf>
1=1
rt Jf> r-i
1\ Bm
l!:f *
mI3 ff J?>
IB *7

Notes

Some of the signs which occur both in Text 1 and in Text 2 differ slightly from ea
other. In Text 2, the -sign and the ke4-sign differ only in their length. In Text 1 they we
of the same length, but differed in the position and length of the verticals. Strictly speakir
the sign-shapes in Text 2 are more "correct".
In line 6, the word Ki-en-gi is divided into two lines within the one case.
50 Lesson 2

Transliteration Transcription Translation

1: dlnanna Inanna For Inanna,
2: nin--ni nin.ani.(r) his lady
3 : Ur_dNammu Urnammu Ur-Nammu,
4: nitab-kalag-g nitabkalaga the mighty man,
I ki-ma
I Unm
5 : .J!gru- lugal.Urim.a(k) the king of Ur,
5
6: lugal-Ki-en-gi-Ki-uri-ke4 lugal.KiengLKiuri.k.e the king of S umer and Akkad

7: --ni e.anL0 her temple -

8 : mu-na-du mu.na.(n.)du.0 he built.
Commentary
2. nin--ni = nin.anL (r), as in Text 1 . Here the nominal phrase expressing the benefactive
consists of the DN and an appositive, which itself consists of a noun with a possessive
suffix.
In this text, nin is used to refer to a goddess. This is the more normal practice; in Text
1 , nin referred to a god.
4. nitab-kalag-g = nitab.kalaga. nitab is one of several S umerian words meaning approx
imately "man".
kalag-gi!, representing Ikalaga/, is an adjective meaning "mighty". Many adjectives in
Sumerian end in la/, representing a morpheme .a. This .a has many uses: formation of
adjectives from verbal roots; nominalization of verbal phrases; marking of certain kinds of
subordinate clauses; etc. It is sometimes called a "nominalizer" or "nominalizing particle"
(although such terms do not cover all its uses). For convenience sake, the term
"nominalizer" will be used here. In this particular case, the adjective Ikalagal is formed
from the verbal root Ikalag/, by the addition of the nominalizer I al. Since the nominalizer
.a is "built into" the adjective, it is not separated-out in transcription. That is, it is
transcribed as: kalaga, and not as: kalag.a. This is further discussed in Lesson 6.
Some Sumerologists prefer to say that Sumerian has no real (morphological) class of
adjectives, but instead has two kinds of "participles", one of which ends in the nominalizer
la/. For convenience sake, however, the traditional term adjective is retained here.
The two cuneiform signs of the adjective are here transliterated as kalag-gi!. However,
the same two signs of this word are often transliterated as kala-gf!.. If one looks in the stan
dard sign-lists for this particular sign (ID), it is given the values kal, kala, kalag, kalag,
and even kal.
Probably all Sumerologists would say that the word for "mighty" was formed from
two morphemes: the root Ikalagl with the addition of the nominalizer la/. They would
also say that the word was probably pronounced something like Ikalaga/. (There are late,
syllabic spellings of this word as kal-Ia-g, etc.) But exactly how do the two written signs
convey this information? There have been three approaches to the problem. One view is to
see the first sign as representing the entire word Ikalaga/. In this case, the following -
 Lesson 2 SI

sign would be a kind of "phonetic complement": It gives some extra information to the
reader, helping him to choose the correct reading of the previous sign. The word might
then be transliterated as: kalagga.
A second view attempts to make the signs approach the transcription. Since this word
is pronounced Ikalaga/, and since the Igal is expressed by the g-sign, this view says that
the first sign must therefore be read Ikala/: kala-g. Thus, this view really derives the
transliteration from the transcription.
The third view says that the transliteration should not necessarily be expected to fit the
transcription. Rather, there are certain general rules of Sumerian orthography which are
found in several different contexts. In this particular case, the transliteration kalag-g!!.
reflects the orthographic rule that a consonant is graphically reduplicated before a word
final (occasionally syllable-final) vowel, particularly across a morpheme boundary. For
example, in Text 1 there occurred: nin-an-na, for nin.an.a(k).
The entire problem is not easy to resolve. Several obvious questions come to mind:
How can one know, for instance, that the sign ID can, in fact, be read as kal, or kala, or
kalag, or kalag, or kalg? To what extent are readings "manufactured", to make the trans
literation more closely approximate the transcription? How valid is the general rule of
Sumerian orthography presented above?
In practice, inconsistencies arise in transliteration, because no matter which
transliteration system is followed, the meaning is normally clear. Whether these two signs
are understood as kalagga, kalag-g or as kala-g, everyone would understand the pronun
ciation to be Ikalaga/, and the meaning to be "mighty". (Even here, however, some S ume
rologists would say that the original form */kalagal > Ikalga/. It is true that similar cases
of vocalic loss are attested in Sumerian. However, the [late] syllabic writings of the type
kal-Ia-g!!. would seem to argue against such an interpretation in this particular case.)
Therefore, some Sumerologists prefer not to deal with these problems, unless they are
interested in the writing system per se.
This problem has been discussed at some length, because it is useful to be aware of the
theoretical principles which underpin our understanding of the writing system. This type of
knowledge is also essential if one is to understand borrowings of the Sumerian writing
system, such as, e.g., the writing system used for Eblaite. And, it is important to be
prepared for (and to understand the reasons for) the inconsistencies and variations in
transliteration which are encountered in Sumerological literature.
In general, adjectives in S umerian follow the noun they modify.
5. illgal-Urims ki-ma = lugal.Urim.a(k), "king of Ur". Just as both kalag- and kala-g
are found in transliterations of the same two signs, so also these signs are found trans
literated as Urimski-ma and UriskC ma.
6. lugal-Ki-en-gi-Ki-Uri-ke4 = lugal.KiengLKiuri.k.e., "king of Sumer and Akkad".
Sumerian has a conjunction meaning "and", linking nouns, but it is relatively
uncommon. Instead, Sumerian prefers to conjoin two nouns directly: an-ki "heaven and
earth".
The first element of the genitive phrase is the single noun lugal. The second element is
formed by the two conjoined nouns, KiengLKiuri. The genitive marker .k follows the two
52 Lesson 2

elements. This can be diagrammed as: lugal.[Kiengi.Kiuri] .k. It is possible for either
element of a genitive phrase to be even more complex, consisting of a noun with a
possessive-suffix, an adjective, a relative clause, etc.
Lines 3-6 form a long nominal phrase, ending in the ergative case-marker .e. This
nominal phrase consists of: a personal name (line 3); an appositive consisting of a noun
and an adjective (line 4); a second appositive consisting of a genitive phrase (line 5); and a
third appositive, consisting of a more complex genitive phrase (line 6).

Discussion: structure
It is instructive to compare the structure of Text 1 and Text 2:
Text 1 :
[Nanna, nin.an.a(k), nin.ani] .(r) benefactive
[Urnammu, lugal.Urim.ak] .e agent
[e.ani] .0 patient
mu.na.(n.)du.0 verb

Text 2:
[Inanna, nin.ani] .(r) benefactive
[Urnammu, nitag.kalaga,lugal.Urim.a(k), agent
Iugal.KiengLKi uri.k].e
[e.ani] .0 patient
mu.na.(n.)du.0 verb

The order of the constituents is the same. As mentioned in Lesson 1 , the constituent
order in these inscriptions is quite formulaic. The difference in the two inscriptions is in the
length of the various nominal phrases, and not in the basic structure.

- Brick-stamps

The cuneiform signs in this text are much more "linear" than those of Text 1 . This is
because Text 1 was "handwritten" by a particular scribe. Text 2 was produced by a "brick
stamp". Brick-stamps were used to mass-produce copies of inscriptions. The writing on
them is done in reverse ("mirror-writing"), so that the impression comes out correctly. The
shape of the signs used tends to be linear, although occasionally they can approach the
shape of the handwritten signs. Several brick-stamps have been preserved, although
apparently none from the Ur III period. The following illustrations are of brick-stamps
from the Old Akkadian period:
 Lesson 2 53

-Case relationships

.ra and the other case-markers in Sumerian are variously referred to as "cases", "case
markers", "case-endings", "postpositions", "postfixes", etc. Strictly speaking, these tenns
are not all synonymous, because they do not all refer to the same level of analysis.
The tenn "dative case", for example, refers purely to a grammatical relationship. This
case can be used to indicate several different semantic relationships: indirect object,
benefactive, etc. "Dative case-marker" or "case-ending" refers to the specific fonnal device
which signals this grammatical relationship, that is, the .ra. "Postposition" or "postfix"
means that the case-marker occurs at the end of a nominal phrase. (This contrasts with
English, for example, where "prepositions" occur in front of a nominal phrase.) Thus, in
Text 1 and 2, .ra can be described as a postpositive case-marker of the dative case, used to
express the bepefactive.
Although these tenns are distinct, in practice they are often used somewhat
indiscriminately. This is because it will normally be clear from the context which level of
analysis is being referred to. Similarly, the dimensional-prefixes are sometimes said to
54 Lesson 2

cross-reference the cases, and at other times are said to cross-reference the case-endings.
Strictly speaking, they cross-reference the case relationships which are marked by the case
endings. For ease of exposition, however, it is usually easier to present them as cross
referencing the case-endings themselves.

- Genitives and cases

The genitive does not behave like the (other) cases in Sumerian, and so it is
occasionally referred to as a "genitive marker", instead of as a case. First, a genitive phrase
can be embedded within a nominal phrase, which can then have its own case-marker. That
is, the genitive can be cumulative with respect to the (other) cases. For example, the
genitive can be directly followed by the ergative case-marker .e, as in Text 1 and Text 2.
The (other) cases, however, are not cumulative with respect to each other. If a nominal
phrase has the dative case-marker, for example, it is impossible for it to have any other
case-marker. Second, the genitive is not resumed by any dimensional-prefix. The dative,
for example, is resumed by the dimensional-prefix .na. However, the genitive is not
resumed.
The reason for the difference in behavior is because of the different r6le which the
genitive plays in a sentence. Genitives relate noun phrases to noun phrases. But the (other)
cases relate noun phrases to verb phrases. That is, genitives and cases perform two
different functions. However, "case" is the term most frequently encountered in
Sumerological literature.
Some scholars use the term "adnominal" case to refer to the genitive and to the
equitative (to be mentioned later). Both can be cumulative, and neither is resumed by any
dimensional-prefix.
The ergative and absolute cases pattern together, in that they are the only cases cross
referenced in the immediately pre- and post-verbal root slot. (In some ergative languages,
verbal cross-referencing only occurs with the agent and the patient, and not with any other
case relationship.)
The cases besides the ergative, absolute, genitive, and equitative are referred to as
"adverbial". They include the following; they will be studied in subsequent lessons:
dative; terminative; locative; locative-terminative; comitative; ablative. Some scholars use
the term "oblique" instead of "adverbial"; others use the term "dimensional". The latter is
rather nice, since these cases are the only ones to be cross-referenced by the dimensional
prefixes.
To sum up, the Sumerian cases may be categorized as:

primary (ergative; absolute)

adverbial (dative; terminative; locative; locative-terminative; comitative;
ablative)
adnominal (genitive; equitative)

-Earlier views of genitive

It was Poebel who definitively established the form and function of the S umerian
genitive ( 1 935). Earlier views were quite different. For example, Franois Thureau-
 Lesson 2 55

Dangin saw the genitive in Sumerian as being formed in two different ways: either by
simple "juxtaposition" of two nouns (lyw-uru, "king-city" = "king of the city"), or by an
ending .a of a "general indirect case". He thought that the Ikl which appears when a vowel
follows the genitive marker was "inorganic"; it was a "hiatus-breaker" to avoid a sequence
of two vowels. Poebel effectively destroyed Thureau-Dangin's views, but traces of the
latter are still encountered in some works. Poebel' s work was further elaborated by
Jacobsen (1973).

- Typology

Scholars have pointed out previously that the genitive in Sumerian behaves differently
than the (other) cases. This is typical of agglutinative languages, where the genitive will be
cumul ative with respect to cases.
In most S-O- V languages, genitive constructions are expressed by the sequence
"possessor-possessed" (regardless of the exact morphological devices used). Sumerian,
then, would seem to be atypical, in that the sequence is "possessed-possessor". It will be
seen later that Sumerian also possesses a genitive construction of the type possessor
possessed, but this construction is not as common as the possessed-possessor one.
In Sumerian, modifiers of nouns typically follow their head noun. In this lesson an
adjective follows its head noun, and in Lesson 7 a relative clause follows its head noun. It
is a general characteristic of S-O- V languages for modifiers to follow their head nouns.

- Proper names

Many different etymologies of Ki-en-gi have been proposed, and just as many ex
planations for the derivation of Sumeru from Ki-en-gi. The sheer variety of such expla
nations shows how unsure such attempts are. Some of those proposed by more prominent
Sumerologists incl ude:

Arno Poebel: Ki-en-giCI) is a dialectal form of kalam "land".

Anton Deimel: Ki-en-giCI) = ki.gir "land of the foot", i.e. "stopping place".

Thorkild Jacobsen: Ki-en-giW = ki-Nigir; Nigir ) Nibir ) Nibur ) Nibru, "Nippur".

That is, the term "Nigir" (whatever this may have meant originally) was at first applied
only to the city of Nippur. Later, in the form "ki-Nigir" ("place of Nigir"), it became
generalized to the whole land of Sumer.

Edmond Sollberger: Sumeru is the Emesal form of Ki-en-gi(r), whatever the ety
mology of the latter might be.

Edmund Gordon: Ki-en-gi = ki.gir1 5 ' "noble place".

Many other dubious etymologies have been proposed. They illustrate the fact that
there is really very little evidence to make a positive decision; the data can be made to fit
many different interpretations.
56 Lesson 2

-Titulature

nitab-kalag-g!! is a very old title, attested even with rulers preceding the Akkad dynas
ty. It was also used by Utu-hengal of Uruk. It is difficult to say exactly what an expres
sion like "strong man" or "mighty man" means; Hallo says: "'strong man' (that is, we
might almost say, independent ruler)" ( 1 966: 1 38).
Ur-Nammu was the first Mesopotamian ruler to use the title l!!gal Ki-en-gi Ki-uri. It
was used by his son Shulgi, but not (apparently) by the other rulers of the Ur III Dynasty.
It was used sporadically by later rulers (in both a Sumerian and an Akkadian form), right
down to the Persian period, especially by conquerors of Babylonia (such as Cyrus).
Ur-Nammu first assumed this title about the fourth year of his rule. In the early years
of his reign, the extent of his control was too limited, and his hold too weak, to permit use
of such a grandiose title.
 Lesson 3

This is another building inscription of Ur-Nammu. Like many royal inscriptions. it

exists in several copies. To illustrate the range of variation in the external shape of the
signs, two different copies of this one inscription are reproduced. Text 3a was produced by
a brick-stamp, as was Text 2. Text 3b was hand-written, as was Text 1. Both copies are
from UT.

This brick is on display in the British Museum. The display stand was obviously
made many years ago. It refers to Ur-Nammu as "Ur-Gur", it refers to Nanna as "Nannar",
and it dates the brick to "about B.C. 2500" - about four centuries earlier than today's
chronology.

57
58 Lesson 3

Sign-list and vocabulary

bad city wall, rampart, fortification

Notes

bAd The PSD translates bad as: "wall", "fortification". Its normal Akkadian equivalent is
duru, glossed by the CAD as: " 1 . city wall, fortification wall, 2. inner city wall, 3.
fortress, 4. enclosure of a house".
 Lesson 3 59

Text 3 a

* A) HI n o>
b>Jffl> If 7H
l'::f *
R=l 1=1
<V1:::1 Jfiil P>
60 Lesson 3

Text 3b

n p
" ii> t>
* lif
If
m
t>

Notes

Both inscriptions are inscribed on bricks, yet the signs in Text 3a are much more linear
than the signs in Text 3b. Because of the clearly wedge-shaped form of the signs, Text 3b
looks more like what we are accustomed to think of as cuneiform.
Differences between Text 3a and Text 3b also occur in lines 4 and 7. The various
signs which make up the geographical names are differently divided between the two lines
in each case.
 Lesson 3 61

Transliteration Transcription Translation

1: dNanna Nanna For Nanna,
2: lugal--ni lugal.ani.(r) his king -

3: Ur_dNammu Umammu Ur-Nammu,

4: lugal-Urimski-ma-ke4 lugal.Urim.ak.e the king of Ur -
5: --ni e.ani.0 his temple -
6: mu-na-du mu.na.(n.)du.0 he built;
7: -.JL- ' ki-ma
b ' d Unm bad.Urim.a(k).0 the city wall of Ur -
S
8: mu-na-du mu.na.(n.)du.0 he built.
Commentary
1 . The first six lines are essentially the same as those of Text 1 and Text 2. In line 2,
however, the tenn lugal is used to refer to Nanna, instead of the nin of Text 1 .
7. The next two lines fonn a new sentence, "he built the city wall of Ur". Line 7 i s the
direct object (patient), consisting of a simple genitive phrase.
8. The verbal phrase is exactly as in the previous inscriptions, and as in line 6.
The verb fonn in line 8 contains a dimensional-prefix .na, and a personal-affix (n.).
However, there is no expressed datival phrase, nor agent, in lines 7 or 8. Rather, the logical
benefactive to which the .na refers is Nanna of line 1 , and the logical agent to which the (.n)
refers is Ur_dNammu of line 3.
The syntax of Sumerian beyond the level of the simple sentence has yet to be
adequately studied. However, it often happens that a verbal chain will contain elements
which cross-reference nominal phrases occurring in a previous sentence.

Discussion: structure
The basic structure of this text is:

[Nanna, lugal.ani].(r) benefactive

[Umarnrnu,lugal.Urim.ak] .e agent
[e.ani] .0 patient
mu.na.(n.)du.0 verb
[bad.Urim.a(k)].0 patient
mu.na.(n.)du.0 verb

- Sign fonnation

Some cuneifonn signs are, in origin, combinations of two different signs, one of
which is pictographic, and one of which is a phonetic indicator of some kind. For example,
the bad-sign is a picture of a city wall, with an inscribed bad-sign (); the function of the
inscribed bad-sig'n is to aid the reader in the correct pronunciation (another example occurs
in Lesson 15).
In Lesson 1, it was mentioned that the second element of the Nanna-sign (),which
62 Lesson 3

was interpreted by later Mesopotamian scribes as a ki-sign, may originally have been a na
sign, functioning as some kind of phonetic indicator.

- Co-ordination

In general, independent sentences in Sumerian are coordinated without any

conjunction. In Text 3, no conjunction appears in line 7. In Lesson 1 6 there occurs the use
of a conjunction g, borrowed from Akkadian.

- History

Most building inscriptions refer to only one undertaking (for example, the building of
a single temple), but it is not uncommon to find such inscriptions referring to two closely
related activities.
The wall referred to in this inscription was undoubtedly the wall which surrounded the
city of Ur. Woolley describes it as follows:
The walled city was in shape an irregular oval, measuring about 1 1 30 yards
in length by 750 yards in width, and was surrounded by a wall and rampart.
The rampart was of mud-brick with a steeply sloping outer face . ... Along the
top of this ran the wall proper, built of burnt bricks. ... Of Ur-Nammu's wall
not a trace remained ... just because the defences of Ur had been so strong
the victorious enemy [that is, the later Elamites who sacked the city] had
dismantled them systematically, leaving not one brick upon another
( 1 982: 1 37-8).
Although Ur-Nammu's wall may have originally enclosed the entire city of Ur, the city
rapidly expanded beyond these walls. The original walled city may have comprised no
more than one-fourth or one-fifth of the city in the Isin-Iarsa or Old Babylonian periods.
The situation must have been similar to that of many Middle Eastern cities today, with a
core consisting of an "Old City" (often referred to as a "madina" in Arabic or English), but
with much extension beyond it.
 Lesson 3 63

Text 3 c
supplemen tary

The supplementary texts included here and in subsequent lessons are meant for prac
tice and review. They will nonnally contain no new vocabulary or grammar; any new
features will be explained.
This particular text is another brick.

rt& 'f

rf
 Lesson 4
This inscription was engraved upon a stone bowl. No photograph is available.

Sign-list and vocabulary

t@f a--- Nin-gal Ningal (DN, fern)

MT>--< tU (ti) to live

Tf . . . J1 . ru
.. to dedicate a votive object

Notes

Nin-gal Ningal was the wife of Nanna, and the mother of Inanna. Being the wife of
Nanna, she was especially worshipped in Ur. Her name means "great lady".

65
66 Lesson 4

Text 4

* Er-
1T
P==r 8
:<l FrtlJ>
If ....

6 t
{J
ij la
f-11l1 $.
-j;
Tf )l
 Lesson 4 67

Transliteration Transcription Translation

dNin-gal Ningal For Ningal,
1:
2: nin--ni nin.ani.(r) his lady -

3: Ur_dNammu Urnammu Ur-Nammu,

4: nitab-kalag-g nitab kalaga the mighty man,

5: .J!
I g- Ski-ma
I Unm lugal.Urim.a(k) the king of Ur,

6: lugill-Ki-en-gi-Ki-uri-ke4 lugal.Kiengi.Kiuri.k.e the king of Sumer and Akkad -

7: nam-tU-Ia-ni-e nam. til.anUe for the sake of his life -

8: -mu-na-ru a.mu.na.(n.)ru he dedicated a votive offering.
Commentary

7. Sumerian does not have many processes of word formation. However, it does have a
formative element nam. Prefixed to verbal and nominal roots, it produces what are called
"abstract nouns". For example, l!!gal is "king", nam-l!!gal is "kingship"; tU is "to live",
nam-tU is "life".
M is a case-marker not seen up till now, marking the "terminative" or "directive" case.
It does not have any one exact translation into English, although it generally indicates
"direction towards" or "action towards". Here, the meaning is something like "for the sake
of', or "on behalf of'. This phrase means something like "for his life", that is, "so that the
king will live a long time".
The exact phonemic value of this case-marker is not actually sure. Since the dative
case-marker Iral ) Irl after a vowel, one might expect Iel ) IU after a vowel. Such a
reduction does occasionally happen, although not apparently in this particular expression
(the sign rB does not appear to have any reading in liU). The original value of this
morpheme may have been, in fact, lee/. Conventional S umerological practice is to simply
transliterate it as e. It will be further discussed below.
The two signs transliterated here as: tU-la are also found transliterated as: !i-la. This
is the same problem seen earlier with kalag-g - kala-g, and UrimskCma - Uris ki-ma. In
this particular case, there is the added irritation that the sign l>-'(T is read !i (with no dia
critic) and til (with diacritic) .
8 . ...ffi, "to dedicate a votive offering". This i s a "compound verb", a type of word
formation very common in Sumerian. A compound verb is made up of two elements. The
first element of the compound verb immediately precedes a complete verbal form, including
its prefix chain. The second element functions as the regular Sumerian verbal root. In this
case, the first element is ; then comes a regular verbal form, consisting of a prefix chain
and the verbal root ru; ru is the second element of the compound verb.
The original function of the two elements of a compound verb is often not clear. In
many cases, the first element is (historically, at least) the patient of the verbal root which
forms the second element. For example, g!l... de means "to speak". de means "to pour out",
and g!l means "voice". Historically, then, this particular compound verb means "to pour
68 Lesson 4

out the voice"; that is, g!l is historically the patient of de. In such cases, the first or second
element may appear elsewhere as a free morpheme, functioning in other contexts like any
other noun or verb.
In other instances, the nominal element of the compound verb is in one of the adverbial
cases. For example, "to found" or "to establish" a temple or building of some kind is:
ki...gar. gm: means "to place"; ki means "ground". Here, ki is in the "locative" case (ki.a,
Lesson 6). The historical meaning was "to place (something) on the ground".
In some cases, the relationship between the nominal and the verbal components is not
sure. In this particular case, for example, ru probably means "to send" (judging from its
use in other contexts as a "simple", that is, non-compound verb), but it is not known what
the element means here.
Synchronically, it is not easy to define the term "compound verb". Certain verbs, for
instance, are almost always used with certain patients (cf. g!l... de above). Are these
compound verbs or not? It i difficult to say; it is not easy to produce a rigorous definition
of compound verbs. Compounds are basically identified on semantic criteria. If the
meaning of the compound is more than the sum of its parts - that is, if it is loosely an idiom
- then it is felt as a compound.
In the sign-lists and vocabularies in this book, compound verbs are indicated by the
use of three periods: ...ru.
In line 7, the terminative case-marker e occurs. According to what was said earlier
about the dimensional-prefixes, one might expect this e to be cross-referenced by a
dimensional-prefix. The dimensional-prefix which cross-references the terminative e is i;
it follows the datival dimensional-prefix in the prefix chain. Therefore, one might have ex
pected to find a form such as: -mu-na-i-(!l)-ru.
This verbal form illustrates the basic problem of the dimensional-prefixes. Although
theoretically there is a one-to-one correspondence between case relationships and
dimensional-prefixes, in practice it isn't so. Case relationships are found that are not appa
rently cross-referenced by a dimensional-prefix, and conversely dimensional-prefixes are
found even when no case relationship is apparently present.
This is probably more than just an orthographic problem. That is, it is not simply the
case that the dimensional-prefix is "there", but not written. In all the Ur III votive
inscriptions, the e in expressions of the type nam.til.anUe seems never to be resumed.
The presence or absence of dimensional-prefixes probably depends on semantic
factors at the level of the sentence, and at the level of the discourse, that is, beyond the level
of one single sentence. Certain nominal phrases are less closely bound to the sentence or to
the discourse than others. For example, nam.til.anUe is only loosely bound to the
sentence; it could be omitted, without any great loss of information. Such loosely-bound
phrases (almost idiomatic or formulaic in character) may perhaps not need to be resumed,
while such important constituents as the benefactive phrase would need to be resumed. Not
a great deal is known about the structure of Sumerian at this discourse level.
To sum up, although the problem is still open to discussion, it is more likely that it is a
semantic-syntactic problem, and not an orthographic problem. Therefore, no dimensional
prefix i is indicated in transcription.
 Lesson 4 69

Discussion: structure
The structure of this text is:

[Ningal, nin.ani].(r) benefactive

[Urnammu, nitab.kalaga, lugal. Urim.a(k), agent
lugal.KiengLKiuri.k].e
[nam.til.ani].M purpose
a.mu.na.(n.)ru verb

- Abstracts and concretes

nam is the regular element used to form abstract nouns in Sumerian. nig is regularly
used to form concrete nouns from verbal roots; this is discussed in Lesson 22.

- Phonology of -M

The reason it is difficult to determine the phonetic shape of e is because it seems to

appear in different forms even under identical conditions. For example, the writing u-zu
u ("towards your hand") would justify an interpretation as: u.zu.e, with the loss of the
final vowel, presumably conditioned by the presence of the lu/-vowel before the IU.
However, the form gir-zu-e ("towards your foot"), which would presumably represent
gir.zu.M, with no reduction of the final vowel, is also found. Such variation can occur
within one text, or even in different copies of one Sumerian literary work. For example,
line 72 of the Sumerian composition entitled "Schooldays" has the expression "towards my
hand", u.gu l O .e. In most copies of the text, this is written: u-g!!l O -M. However, at
least one copy has: u-g!!l O-u.
The problem is further compounded by the fact that the e-sign also has a reading e,
so that if the terminative case-marker follows a word ending in lel, the writing is
ambiguous.
It is possible that the writings in -e should be understood as morphographemic,
standing for: I/. The scribe wrote the full, more original form of the morpheme, even
though in certain phonetic contexts it had been reduced in speech. That is, -e is written
conventionally for the terminative case-marker, without regard for its precise phonetic
shape. Other such morphographemic writings occur in Sumerian.

- Usage of -e

As stated above, the terminative case marked in -e in this fixed expression is

apparently not cross-referenced in any of the Ur III royal inscriptions. Curiously enough,
however, there are a few cases in royal inscriptions from earlier periods where the
tenninative in -e in such expressions is so cross-referenced by its dimensional-prefix.
This is strange, because the early texts are usually less explicit than the later texts. Also,
this dimensional-prefix often appears in the writing of Ur III administrative texts. Such a
distribution indicates that factors such as genre and style help determine the use or non-use
of the dimensional-prefix .i (and of the use and non-use of the dimensional-prefixes in
general).
70 Lesson 4

For anyone who has studied Akkadian, the Sumerian terminative in Iel immediately
calls to mind the Akkadian terminative in liU. To some degree, the two morphemes
overlap both in form and in function. At least as far back as 1 925, it was proposed that the
Akkadian terminative morpheme was borrowed from Sumerian (Albert S chott). However,
most scholars do not accept this view, for two reasons: First, the Akkadian terminative
appears to have cognates in other Semitic languages; it has recently turned up in Eblaite, in
certain limited contexts. Second, there do not appear to be any cases where Akkadian has
borrowed Sumerian grammatical morphemes. Similarly, it is difficult to believe that the
Sumerians borrowed the morpheme from the Akkadians. Although rare, it does show up
in very early Sumerian texts.
It is more likely that the two are independent developments. It is not impossible,
however, that the chance formal similarity between the two morphemes has caused the two
to influence each other in meaning, pulling them closer together in meaning than they may
have been at some earlier period. Given the fact that Sumerian and Akkadian were in close
contact for over a millennium, such reciprocal influence upon the grammar is not too
surprising.

- Compound verbs

Most compound verbs are of the type noun-verb, where the noun is (historically) the
patient of the verb. It is not sure whether these nouns are to be regarded as patients in
synchronic terms; this problem is discussed in Lesson 1 2. In some cases, the noun is in
one of the adverbial cases. However, more complicated compound verbs also occur, of
differing types: adjective-verb, noun-adjective-verb, and even noun-noun-verb. In the case
of the latter, one noun is (historically) the patient, and the other is (historically) in an
adverbial case. An example is "to pray", kiri3 .u ... W, literally, "to place (g@ the hand
(]y) on the nose (kiri3 )"; kiri3 is either in the locative case (kiri.a) or the locative-terminative
case (kiri.e).
Because there is no obvious formal way to define compound verbs, it is a legitimate
question to ask whether such a class of words actually exists. If more were known about
the etymology of each individual case, one might be less inclined to even posit the existence
of the class of compound verbs.

- Votive inscriptions

This text is a typical votive inscription, a second subclass of royal inscriptions as

distinguished by Hallo. Votive inscriptions are found on objects which were actually
donated and placed in a temple. The object can be of various kinds; this particular
inscription is found on a stone bowl. Possibly, such vessels would have been used for
ceremonial food offerings. Since stone is not common in Sumer, such a vessel would have,
been considered as something special. In this book, the following votive objects occur: a
stone headdress (Text 1 1 ), a vase (Text 1 2), beads (Texts 1 6, 1 8a, and 1 9a), and a cylinder
seal of limestone (Text 22).
The purpose of such votive objects was to convey a hope from the donor for the long
 Lesson 4 71

life of the king (the donor might or might not b e the king himself). The objects were not
"functional", as we would understand the term; that is, this bowl was not actually used as a
daily eating utensil. Similarly, the votive cylinder-seal in Text 22 was probably not used as
a daily, routine, cylinder-seal.
Except for the component of the compound verb, there is no direct object (patient) in
the text. This is because the votive object itself can be thought of as constituting the direct
object (patient).
The use of the term "votive" to describe such inscriptions has been criticized by A.
Grayson:
The etymology of the word "votive" implies a vow and, since no vow is
involved in the ancient Mesopotamian texts under discussion, the term is
incorrect. They are certainly not "votive" or "ex-voto" inscriptions in the
ancient Roman sense where a vow preceded the dedication ( 1980: 1 57).
For this reason, Grayson prefers to refer to such texts as "dedicatory inscriptions".
72 Lesson 4

Text 4a
supplemen tary

This is another brick.

Sign-list and vocabulary

q Utu Utu (ON, masc)

Notes
Utu This was the son of Nanna, and older brother of Inanna. He was primarily the god of
the sun (his name means "the [visible] sun"), but was also connected with truth, justice, and
law-giving. He was equated with the Semitic Shamash, who is pictured on the top of the
stela of Hammurapi as handing over the law code written on the stela to Hammurapi. Utu
was especially worshipped in Larsa and Sippar.
 Lesson 5 73

Lesson 5

This text is another building inscription, inscribed on a clay cone.

Sign-list and vocabulary

En-lil Enlil (DN, masc)

":y En-erin2-nun Enerinnun (canal name)

1< kur mountain; highland; foreign land

i d (17) river, canal

* A1 nidba food offering

ba-al to dredge, excavate

::3"
ra

t::f ka

Notes

En-IU Father of Nanna; the most important god of the Sumerian pantheon. His name
means "Lord Air" or "Lord Wind", but he was in general responsible for the orderly
running of the universe (although he had a destructive side as well).
Enlil functions as the active leader of the Sumerian gods; he has apparently displaced
the sky-god An from this role. He was worshipped at many places, but his special
sanctury was the E.-kur in Nippur.
The lil-sign is the same ke4 -sign seen previously.
(The term en, "lord", is discussed in Lesson 9.)

En-erin2 -nun The location is uncertain, as is the etymology (although it is presumably

Sumerian).

k:ur The original meaning of this word was probably "mountain"; the kur-sign, in fact, is
thought to be the picture of three mountain tops. The word then comes to mean "foreign
land".

id This sign is composed of two elements: If

, which by itself represents , "water", and
ID which by itself represents ur, "watery deep, sweet waters" (see Lessons 1 and
14) , or Nammu the goddess. The original meaning of the sign may then have been
74 Lesson S

something like "water coming from the sweet-waters", as opposed to "water coming from
the salt-waters", and as opposed to "rain" (Mg, which is composed of the -sign followed
by the an-sign: TT ).

nidba The reading is uncertain; it is also transliterated as nindaba. When scholars are
unsure of the reading of a "compound logogram" (a single "word" graphically composed of
several individual logograms), they occasionally add in parentheses the reading of the
component parts. Thus, this sign is also transliterated as: nidba (PAD-dINANNA), or any
of several variants, such as: nidba (SUKUR2 _ dINANNA), since it is in fact not clear
exactly what all the components of this particular logogram are!
This word was borrowed into Akkadian as nindabu, although the forms nindabbu,
nigdabbu. nidabu. nidbu and nidpu also occur. CAD translates the Akkadian word as:
"cereal offering, food offering, provisions".

ba-al This is the normal verb used to describe the restoration of a canal, that is, clearing it
by dredging it of accumulated silt and trash.
The verb is almost always written in this way, with two signs. There are a few
instances where it is spelled bal or ba-la. It is not sure what the writing ba-al implies about
Sumerian phonetics. The vowel may have been long, or there may have been a glottal stop
or a glide between the two / a /-quality vowels. Because of this unusual writing, it has even
been speculated that the word is a borrowing from an as-yet unidentified language. For
convenience sake, it will be transcribed here as: ba-al.
 Lesson 5 75

Text 5

Jb4
p ,
f-J( !l
$.t
\1ijg W
1 ii
yY Jf \ p.- -r
Yf ttr ,O
rw
76 Lesson 5

Transliteration Transcription Translation

1: dEn-lil Enlil For Enlil,
2: lugal-kur-kur-ra lugal.kur.kur.a(k) king of the lands,
3: luw--ni lugal.anL(r) his king -
4: Ur_ dNammu Urnammu Ur-Nammu,
5: ' ki -ma
1 !giL1-Unm
...! lugal.Urim.a(k) the king of Ur,
S
6: luw-Ki-en-gi-Ki-uri-ke4 lugal.KiengLKiuri.k.e the king of Sumer and Akkad -
7: --ni e.anL0 his temple -
8: mu-na-du mu.na.(n.)du.0 he built;
9: i d En-erin2-nun id Enerinnun the Enerinnun canal,
10: i d-nidba-ka-ni id.nidba.k.ani.0 his canal (productive) of food -
1 1 : mu-na-ba-al mu.na.(n.)ba-a1.0 he dredged.
Commentary
2. There are two common plural formations of nouns in Sumerian. For animate nouns, the
plural is formed by a suffixed .ene. For example, "gods" is digir.ene, written Qigir-re-ne.
For inanimate nouns, the plural is formed by reduplication of the noun: kur-kur, "lands"
(This type of plural formation is also occasionally used for animate nouns.)
It is quite probable that such reduplicated plurals were phonetically reduced, but
because of the morpheme-bound nature of the script, such reduction does not normally
show up in the writing.
In Text 6, the plural noun is the second element of a genitive phrase:
lugal. [kur.kur] .a(k), "king of the lands: '.
9. id means "canal" or "river". The name of the canal forms an appositive.
10. A second appositive. The suffix .ani refers to the entire genitive phrase: "his [canal of
food offering]". The genitive phrase "the canal of his food offering" would be:
id. [nidba.ani].a(k). The sequence of third-person possessive-suffix followed by the geni
tive marker is /anak/, not the expected /anik/, so this would be written id-nidba-na.

Discussion: structure
The structure of this text is:

[Enlil, lugal.kur.kur.a(k), lugal.ani] .(r) benefactive

[Urnammu, lugal.Urim.a(k), agent
1ugal.KiengLKiuri.k].e
[e.ani].0 patient
mu.na.(n.)du.0 verb
[id Enerinnun, id.nidba.k.ani] .0 patient
mu.na.(n.)ba-a1.0 verb
 Lesson 5 77

- Writing system
The word for "food offering" illustrates a problem not yet seen in the script. Its
pronunciation as Inidbal is given by various lexical lists, where it is spelled out syllabically
as: ni-id-ba. It is thought to derive from ninda "bread, food" and ba "to divide, to ap
portion"; ninda would presumably be an incorporated direct object (patient) of ba. (It is
occasionally spelled ninda-ba.) The pronunciation as Inidba/ shows a phonetic reduction
of */nindabal > Inidba/. The various spellings in Akkadian, mentioned above, reflect both
the older and later S umerian pronunciations.
It is difficult, however, to say how this word came to be written as "PAD
dINANNA". That is, there does not appear to be any way to phonetically relate the word
Inindabal or Inidbal to the individual signs forming this logogram. Rather, the reading
Inidbal refers to the "sum" of the three signs. Without the evidence of lexical lists, in fact,
there would probably be no way to figure out that this group of three signs was to be read
as Inindabal or Inidba/. The pad-sign has several different readings, and in Akkadian
stands for several different words: kusapu, "a kind of bread" (probably of Semitic
etymology); kurummatu, "food portion" (a Sumerian word), etc. The significance of the
Inanna-sign (if that is how it is to be understood here) is uncertain.

- Proper names

In the bilingual texts from Ebla, the equivalent of En-lil is given as: I-li-Iu. This
seems to agree with later Akkadian pronunciations of the name, which also show an
assimilation of IEnlill > IIllil/; some Sumerologists, in fact, transliterate the two signs En
lil together as: Elli!. (It has also been speculated that the interpretation of the name as
"Lord Wind" is a Sumerian folk-etymology, and that the word is of pre-Sumerian
etymology.)

- Cones

The building inscriptions seen up to this point have all been inscribed on bricks. Text
5, on the other hand, was inscribed on what is commonly known as a "clay cone" or "clay
nail". Clay cones were used throughout Mesopotamian history; their form and function
varied to some degree from period to period. A detailed description is that of Grayson,
describing the clay cones of the Neo-Assyrian period. The clay cone
is an oblong conical object of clay. It is tapered almost to a point at one end
and at the other there is a large semi-spherical head. The same inscription
usually appears on both the shaft and head. The shaft was commonly
inserted in the upper portions of walls with the head, which was painted a
bright colour, protruding ( 1 980: 145).
In this view, the clay cone would have been at least partially visible to on-Iookers.
Other scholars belive, however, that the protruding end would have been plastered over,
covering up the inscription (at least, in the Ur III period). In fact, the exact purpose of these
cones is still something of a mystery to us.
Woolley found such cones in situ, forming part of the terrace of the ziggurat of Ur-
78 Lesson 5

Nammu:
Such cones were familiar enough as objects on museum shelves, but now for
the first time we saw them in position just as the builders had set them four
thousand years before. ... One felt a quile unscientific thrill at seeing those
ordered rows of cream-coloured knobs which even the people of Ur had not
seen when once the tenaced wall was finished and plastered (1982: 140).
The latest such cones found in Ur date from the Neo-Assyrian period. Curiously, they
were not in the wal1, but were buried below the floor.
Clay cones were usually inscribed in several duplicates; at least eight copics ofTcxt 5
are known. The point of such cones, again quoting Woolley, was not to "parade [the
ruler's] achievements before his fellow-men, but to keep the record of his piety fresh in the
mind of the god, who presumably can see through a brick wall" (1982:228).
Some scholars differentiate between clay "nails" and clay "cones". Gelb says that clay
nails "are easily recognized by their mushroom shape. with broad, thick heads and short
shafts. ... Clay cones are characterized by a total or almost total lack of the head"
(1948:267). Different places and periods seem to prefer one or the mher; it is also possible
they had different functions. Gelb says that "the function of such nails and cones is much
the same as that of tablets commemorating the erection of public SlI"Uctures in modem
times" (1948:268).
No photograph of Text 5 is available. The fOllowing is a photograph of a dedicatory
cone from the rime of Gudea's father:

It was mentioned in Lesson 3 that it is not too uncommon for building inscriptions to
treat more than one activity. In the case of building inscriptions inscribed on clay cones, it
is less common for more than one acriyity to be mentioned. In this particular text, the
Enerinnun canal may have brought the waters which irrigated the fields of the temple being
rebuilt.
All of the Ur HI kings were involved with repairing the canals and drainage systems
of Mesopotamia, by dredging and rebuilding. The names of many canals in use during the
Ur ID period are known; most of the names are Sumerian. Urnammu is known to have
built and to have repaired a number of canals. It is usually assumed that much damage had
been done to these systems by the Guti, who are thought to be responsible for bringing
about the fall of the Dynasty of Akkad. Revisionist thinking, however, believes that the
Guti did not do as much damage as is commonly thought, nor did they hold that much
control over Mesopotamia. Most scholars seem to feel that inscriptions such as Text 5
refer more to routine maintenance and expansion of the canals; various kinds of
administrative texts refer to such activity.
 Lesson 6
This is another brick of Ur-Nammu.

Sign-list and vocabulary

t:=r Determinative preceding objects of wood. Transliterated by a superscript "gi".

An An (DN, masc)

%- digir god

t=1 1irJ>--r kiri6 garden

1- barag (bar) dais

ki place, earth

!L=T(ry man to be splendid, magnificent

0Dr sikil to be pure, clean

t:=J gub to stand; to make stand, to plant

Notes
gi In addition to its use as a determinative, gi means "tree; wood; object made of wood".
The Akkadian equivalent, i1!, is translated by the CAD as: " 1 . tree, 2. timber, lumber,
wood, wooden implements, aromatic wood, firewood, 3. wooded area". It is sometimes
transliterated as ge.

An The god of the sky. At one time he may have been the active leader of the S umerian
gods, but at some point prior to our written records he was displaced in this role by Enlil.
Scholars sometimes refer to him as "shadowy", or as a kind of deus otiosus.
His name is almost always written without the divine determinative.

digir It is usually assumed nowadays that this word was pronounced with a velar nasal.
Some think it may have been pronounced I dingir/, and in fact it is most commonly
transliterated as dingir.

79
80 Lesson 6

gBkiri6 This is also transliterated with an initial / g/, especially in older transliterations:
giri 1 1 and giri 1 2 . Some Sumerologists do not think that the gi-component here is a
determinative, and so it is also transliterated: gi-kiri6 . (This is also the view found in
some lexical texts.) However, other Sumerologists believe that the kiri 6 -sign includes what
here is called a determinative; that is, the one sign kiri6 is composed of two separate ele
ments: Cl' and.mr:J , and so there is no determinative. The most current practice is to
read the two signs as gikiri 6 .

barag The PSD translates this simply as "dais"; many Sumerologists translate it as
"throne-dais". It was borrowed into Akkadian as parakku , which is translated by AHw as:
"Kultsockel, Heiligtum".
Text 6 was inscribed on a brick which apparently was part of the barag of the temple.

Id It is not known what the cuneiform sign is a picture of.

mal} This is not common in finite verbal forms. The most common Akkadian equivalent
of its use as an adjective is Trl1, translated by the CAD as: "first-rank (in importance,
quality), outstanding (in size), august, excellent (used only as a poetic term)".

sikil This is also not common in finite verbal forms, but its adjective is frequent. The
usual Akkadian equivalent of the adjective is ellu, glossed by the CAD as: " 1 . clean, pure,
2. holy, sacred, 3. free, noble".
 Lesson 6 81

Text 6
82 Lesson 6

Transliteration Transcription Translation

1 : An lugal-digir-re-ne An lugal.digir.ene.(k) For An, the king of the gods,
2: lugal--ni lugal.ani.(r) his king -
3: Ur_dNammu Urnammu Ur-Nammu,
4: lugill-Urimski-ma-ke4 lugal.Urim.ak.e the king of Ur -
5 : gikirio-
-
mah kiri.mab 0 a magnificent garden -
6: mu-na-gub mu.na.(n.)gub.0 heplanted;
7: barag ki-sikil-la barag.0 ki.sikil.a A dais, - in a pure place -
8 : mu-na-du mu.na.(n.)du.0 he built.
Commentary
1. The sign can be very confusing in context. It can represent the detenninative before
divine names; the god An (as in Text 6); or the word digIT, "god" (or the word an, "sky") .
. ene is the nonnal plural marker for animate nouns. In the previous lesson, re
duplication occurred as the nonnal plural marker for inanimate nouns.
5. mab is an adjective from a verbal root. As discussed in Lesson 2, some of what are
translated as adjectives end in the nominalizer lal, e.g., kalag-g = kalag.a = Ikalaga/.
Other adjectives do not. Certain adjectives sometimes appear with the nominalizer lal, and
sometimes without it, but with no apparent difference in meaning.
The fonns in .a are sometimes described as "passive participles", and those in .0 as
"active participles". For example, the adjective mab could be described as an "active
participle in .0", mab.0. This, however, may not be a valid use of the term "participle".
Although etymologically gikiri6 -mab is composed of two words, it may have been
"felt" as one word. It was borrowed into Akkadian as kirimabg, translated by the CAD as
"pleasure garden".
7. The I gl of barag is an amissable consonant, so the sign has both a long value (barag)
and a short value (bara2 or bara). Since in Text 6 Igl is word-final, some Sumerologists
would transliterate it here by bara2 .
sikil is another adjective from a verbal root. It is like mab, in that it appears without
the nominalizer I al.
The lal which fonns part of the la-sign is a new case-marker, that of the "locative"
case. This case is used to express location, either spatial ("in that place") or temporal ("on
that day"). Spatially, it can usually be translated by the English "in".
Its basic fonn is la/. Following a vowel, it usually does not contract. "In the earth",
for example, is almost always written: ki-. After a consonant, it is nonnally written by a
sign which reduplicates the final consonant of the previous word. Thus, sikil.a is written
sikil-la.
8. The dimensional-prefix which cross-references the locative is .ni. In a sequence of
dimensional-prefixes, it always occurs last. In this line, one might have expected to find:
mu-na-ni-(n)-du, but the dimensional-prefix does not appear. In fact, in the Ur III royal
 Lesson 6 83

inscriptions, the locative .a is usually not cross-referenced. This is probably more than just
an orthographic problem. In Lesson 4, an instance occurred where the dimensional-prefix

for the terminative was not expressed. There, it was said that it may have been because of
the idiomatic nature of the phrase. However, such an explanation hardly fits the numerous
cases where the locative case is not cross-referenced. Perhaps locative phrases in general
were felt as less closely bound to the verbal phrase than were the other adverbial cases.
The logical antecedent of the datival dimensional-prefix .na is the nominal phrase in
lines 1 -2 of this inscription, which is part of a different, independent sentence. The same
situation occured in Text 3 , where a datival dimensional-prefix referred logically to a
nominal phrase in a preceding sentence.

Discussion: structure
The structure of this text is:

[An, lugal.digir.ene.(k), lugal.ani] .(r) benefactive

[Umammu, lugal.Urim.ak] .e agent
[kiri.mab] 0 patient
mu.na.(n.)gub.0 verb
[kLsikil] .a place
mu.na.(n.)du.0 verb

-Amissability

It may be useful here to summarize the ways the amissable consonants are reflected in
the writing system. In the case of grammatical morphemes (which are normally written
syllabically), such as the genitive .ak, the amissable /k/ does not show up in word-final
l
position: "King of Ur" is written: lugal-Urim t ma. When not in word-final position, it
does show up: The same expression, with an ergative case-marker, is written: l!Jgal-
Unm ki-ma-cl4.k
S
In the case of lexical morphemes (which are normally written logographically), it
cannot be determined from the writing system whether the Auslaut was pronounced or not.
That is, just by looking at the sign, there is no way to tell whether the word for "dais" was
read /barag/ or /bara/.
When the amissable consonant of a lexical morpheme is not word-final, there is less of
a problem in understanding the phonology, but still a problem in understanding the nature
of the orthography. As discussed in Lessons 2 and 4, should a form such as "for the sake
of his life" be understood as nam-ti I-la-ni-e or as nam-!l.-la-ni-se? Or should the adjective
for "mighty" be understood as kalag-g.a or as kala-gf!? Phonetically, probably all Sume
rologists would understand these forms to be /namtilange/ and /kalaga/. The question,
rather, is how does Sumerian orthography represent these pronunciations.

- Loan words

In a number of early loan words from Sumerian into Akkadian, final voiced stops
appear as unvoiced, and usually as geminated: isib ("kind of priest", Lesson 19) ) mppu;
Mrig ("an official") ) abarakku. Word-initial voiced stops usually appear as unvoiced in
84 Lesson 6

Akkadian: barag ) parakku.

It is not sure what this tells us about Sumerian phonology. As was discussed above,
the difference between the two sets of consonants Ib d gl - Ip t kl may not have been one
of voice.

-Locative

Locative phrases such as "in the earth", ki.a, are almost always written ki-. That is,
the lal of the locative case-marker almost always appears in the writing. It has been
speculated that the lal of the locative case-marker actually assimilates into or contracts into
a preceding vowel, but the script is morphographemic, and writes the lal anyway.
The locative case-marker .a does not usually appear when the head noun is the first
element of a compound verb (kiri 3 .u ... gal was mentioned in Lesson 4). It is possible that
this is more than a case of assimilation or contraction. As discussed in Lesson 4, it is not
sure how "present" case-markers were in the case of compound verbs, in a synchronic
sense; they may have been deleted or somehow reinterpreted.
Although in the Ur III royal inscriptions the locative case is usually not cross
referenced by a dimensional-prefix, in the contemporaneous Ur III administrative docu
ments it quite frequently is cross-referenced. As was mentioned in Lesson 4, this may
mean that factors such as genre and style must be taken into account, to describe and
explain the different distributions.

-Adjectives

In Text 6 the adjectives man and sikil occur. In Text 2 the adjective kalag-g occurred.
As stated above, some adjectives end in .0, others in .a. Some adjectives are occasionally
found sometimes in .0, other times in .a. This situation is not well understood. The most
recent discussion is by Joachim Krecher ( 1 978). He believes that, at least in certain cases,
the forms in .a mark a nominal phrase as "definite" or "determined" in some way; those in
.0 are the unmarked forms.
The difficulty in investigating this problem (and many other problems in Sumerian
grammar) is that it is not easy to find sentences which are very close in structure, but differ
only in the presence or absence of .a on an adjective. There are usually too many variables
involved, to be able to sort them all out.
 Lesson 7
Two copies of this inscription are reproduced. The variation in the shape of the
cuneifonn signs is fairly minimal. Text 7a is a stone foundation tablet. Text 7b is a brick.
The inscription appearing on them is a standard inscription, a further sub-class of royal
inscription as distinguished by Hallo.

Sign-list and vocabulary

1t1 man

Notes

Id According to Gelb, "The S umerian word 1t1 is a noun meaning 'person, man' (in the
sense of homo, Mensch, not vir, Mann), and may be used for both males and females"
( 1 979b:5 1). Jacobsen says that 1t1 "denotes a man (Akkadian awIlum) or woman (Ak
kadian aWIltum) who heads a household, firm, or city" ( 1 987: 1 30 n. 17).

85
86 Less<!n 7

Text 7a
 Lesson 7 87

Text 7b
88 Lesson 7

Transliteration Transcription Translation

1 : Ur_dNammu Urnammu Ur-Nammu,
2 : ....!
1 !g.fL- Ski-ma
l Unm lugaI.Urim.a(k) the king of Ur,
3 : lugal-Ki-en-gi-Ki-uri lugal.Kiengi.Kiuri.(k) the king of Sumer and Akkad,
4: lu _ dEn-1il-la lu e.Enlil.a(k).0 the man who built the temple
of EnliI.
5: in-du- Ln.du.0.a
Commentary
4. _dEn-lil-la = e.Enlil.a(k). Following certain words which end in Ill, the la-sign (in
stead of the la-sign) is used to express the combination of the reduplication of the previous
word-final III with the lal of the genitive marker, or the lal of the locative case-marker.
Such a word is liI. There are other words ending in III where the la-sign (and not the la
sign) is used. In Lesson 6, there occured: ki-sikil-la. These differences in writing are a
clue that Sumerian had more than one type of Ill-sound. Not enough is yet known about
Sumerian phonology to say exactly what this means.
Lines 4-5 correspond to a relative clause in English, "the man who built the temple of
Enlil". In Sumerian, relative clauses (for want of a better term) are composed of two
elements: a noun serving as relative marker (corresponding in function to an English
relative pronoun), and a nominalized sentence, which stands in apposition to the preceding
relative marker. In Text 7, the relative marker is lu. The nominalized sentence is formed by
the addition of the nominalizer .a to a complete sentence. For example, "he built the temple
of Enlil" is: e.Enlil.a(k).0 mu.(n.)du.0. This can be nominalized by the addition of .a:
[e.EnliI.a(k).0 mu.(n.)du.0].a, meaning something like "(the one) who built the temple of
Enlil". Put into apposition with the relative marker lu, this means "the man who built the
temple ofEnlil".
The simplest way to understand relative clauses in Sumerian is to think of them as
being equivalent to big adjectives. For example, "the mighty man" is: [nitabHkalag].a.
"The man who built the temple of Enlil" is: [luHe.EnliI.a(k).0 mu.(n.)du.0].a. (As a
matter of fact, many linguists consider adjectives in general to be "reduced" relative clauses:
"the mighty man", in English, derives, in some sense, from "the man who is mighty".) The
principle is the same; adjectives and relative clauses fulfill the same function.
Sumerian does not have a morphologically distinct class of "relative pronouns".
Because of the frequency of the noun lu in these contexts, however, it functions almost like
a relative pronoun. The Akkadians themselves sometimes translated lu as a, the Akkadian
relative pronoun. Occasionally, other words may function as the equivalent of a relative
pronoun.
Although such a construction as: lu e.EnliI.a(k).0 mu.(n.)du.0.a is theoretically
possible, in fact the verb form used in Text 7 is different in two ways from the rather stock
or formulaic verbs seen in the previous texts. First, it uses a different conjugation-prefix.
The conjugation-prefix which occurred in all the previous inscriptions was Imul, written
mu. However, this text uses the conjugation-prefix lit. If not immediately preceded by a
 Lesson 7 89

consonant, this conjugation-prefix is nonnally written by the i-sign (that is, the sign which
occurred earlier with the value ni).
Second, here the personal-affix 1nl cross-referencing the agent is expressed in the
script. The in-sign expresses the combination of the .i conjugation-prefix with the .n of the
person al-affix. This particular use of the in-sign is quite regular.
It is difficult to say why the finite verb fonns seen previously use the conjugation
prefix !lli!, but the nominalized fonn uses the conjugation-prefix i. As discussed in Lesson
1, the essential difference between mu and i is elusive. Jacobsen says that the conjugation
prefix i "presents the occurrence denoted by the verb as touching on the subject without
inwardly conditioning him in any lasting manner" ( 1 965:25 1). mu is the
mark of location of the occurrence denoted by the verb on the inside border
(.u) of the area of the speech situation (m.) . ... It adds to this implications of
emotional involvement of the speaker, of his being personally engaged
( 1965:254).
However, not all scholars are as certain as Jacobsen in their conviction; Poebel, for
instance, seemed to have the opposite view of the relationship between mu and i. J.N.
Postgate has expressed perhaps the most negative view:
For many years a vexed question in Sumerian has been the distinction
between the prefixes mu- and i-, and our failure to define the difference in a
satisfactory way has epitomized our helplessness before Sumerian grammar
as a whole ( 1974:24).
And, as discussed in Lesson 1, in actual practice the conjugation-prefixes are basically
ignored in translation.
It is likewise difficult to say why the personal-affix 1nl is written here, whereas it did
not appear in any of the previous finite verbal forms. In the previous texts, the finite verb
fonn always had an expressed subject (agent). In this particular text, there is no expressed
agent, since the verb is inside a relative clause; the h.i of line 4 is the relative marker, not an
agent of the sentence as a whole. One might hypothesize a rule such that "expressed agents
do not use the personal-affix 1nl, but verbal forms without expressed agents do", but from
other texts it is known that the situation is not nearly as simple as this.
To sum up, III of line 4 is the relative marker of the relative clause. The relative clause
is fonned by nominalizing a finite sentence, by use of the nominalizer .a. The nominalized
clause stands in apposition to the relative marker.

Discussion: orthography
In both copies of this text, the relative clause is split into two cases. The first case
contains the more nominal component, and the second contains the more verbal component.
It is not uncommon for long relative clauses to be split into two or even more cases.
The S umerian word for "king", illgal, is a noun-adjective compound from Ill, "man",
and gal, "great, big". In older forms of the illgal-sign, the gal-component was written a
little above and to the right of the Ill-component. At times the two signs were totally
separated, and can even be written on two different lines of one case. As the cuneifonn
signs gradually became more linear, the gal-sign shifted position, and so in "standard"
90 Lesson 7

Sumerian the illgal-sign is all one sign, with the gal-component in front of the hl
component.
As will be discussed in Lesson 1 3, the order of cuneiform signs within a case in the
earliest texts was to some degree free, with the order-as-written not necessarily reflecting
the order-as-read.

- Conjugation-prefixes

As hinted at above, the fact that the conjugation-prefix 1. appears in the relative clause
instead of mu raises several obvious questions: Does 1. appear outside of relative clauses?
Does mu appear inside of relative clauses? What about the distribution of 1. and mu in
general? What about different types of relative clauses: clauses where the relative clause
modifies an agent ("the man who built"), clauses where the relative clause modifies a
patient ("the house which the man built"), etc. What about the distribution of the
conj ugation-prefixes in topic ali zed and emphatic sen tences?
Unfortunately, not all of these questions can be answered. The data are both
ambiguous and limited; the number of attested relative clauses is not that large. Without
access to native speakers, such questions cannot be tested.
In the Ur III texts, the verb form in the relative clause in this particular expression
seems always to appear as: in-du-. In some bricks of Gudea, the form mu-na-du is used
in a sentence as a main verb, while the verb of an embedded relative clause uses the form
in-du-:

dNin-gi-zid-da digiI-ra-ni
Gu-de- ensi2 Laki III E-ninnu dNin-gir-su-ka in-du-
Gir-suki-ka-ni
mu-na-du.

For Ningishzidda, his god -

Gudea, the ensi of Lagash, the man who built the Eninnu of Ningirsu -
built his Girsu temple.

The most recent investigation of the conjugation-prefix lil is by Herman Vanstiphout

( 1 985). He believes that the oppostion mu vs. i cannot be understood simply on the level
of the individual sentence; rather, the larger context (discourse) must be examined. His
tentative results are that
Ii/ seems to carry substantially 'secondary' information in discourse
(accompanying or descriptive information, including consecutive verbs) ...
on the supra-sentential or discourse level the prefix Ii/ serves as a back
grounding device ( 1 985: 1 1 ; 1 3).
He does not examine relative clauses specifically, but such an investigation might prove
useful.
It has also been suggested that 1. was a nasalized vowel, and that the presence or
absence of a following !! is a question of phonology, not of morphology or syntax. But it
is not even sure that nasalized vowels existed in Sumerian.
 Lesson 7 91

- Standard inscriptions

It will be noticed that there is no finite verbal fonn (or any other predicate) in this text!
Instead, there is only a personal name, followed by a series of titles or epithets. Jacobsen
has referred to such sentences as "label sentences". Hallo, in his discussion of the different
categories of royal inscriptions, refers to such inscriptions as standard inscriptions or
property inscriptions. By "standard", Hallo means that the text is something like a flag or
other identifying device: It identifies the building as being the property ofUr-Nammu.
Such inscriptions, which serve to indicate ownership of a building, are often regarded
as the "simplest" fonn of royal inscriptions. They usually consist of a royal name,
followed by a limited number of epithets, one of which may be a relative clause, as in Text
7a/b. There is no verbal predicate, These texts can be very short; Text 1 3c is a standard
inscription consisting of only two lines.

- Foundation deposits

The building inscriptions presented here up to now have been inscribed on either
bricks or clay cones. Text 7b is such a brick. A third category of building inscriptions is
referred to as "foundation deposits". These were actually buried under the foundations of
walls in a building, in a small pit. Text 7a is one sub-type of such foundation deposit, a
stone foundation tablet. In Woolley's words:
Foundation-deposits are found in the corners of buildings. Built into the
wall-foundations there is a small box of burnt bricks, lined with matting and
waterproofed with bitumen; in it is set a copper figure of the king modestly
represented as a labourer carrying on his head a basket of mortar; at his feet
is a stone tablet in the fann of a pIano-convex brick; on the brick and on the
king' s skirt is an inscription recording his name and that of the temple
( 1982: 1 61).
Occasionally, the figurine or the tablet is uninscribed. It is thought that these
foundation tablets were intended to represent "model bricks".
The following photograph is of a figurine which was found in Nippur along with Text
7a, and so bearing the same inscription. (At least two of these bronze canephore figurines
are known.) The figurine is thought to represent Ur-Nammu himself, carrying a basket on
his head with the building materials used to make the "first brick" of a building.
 r.esson 7

92
 Lesson 7 93

Text 7c
supplementary

Another brick.
 Lesson 8
This text was inscribed on what is commonly called either a door socket or pivot
stone.

Sign-list and vocabulary

tHH i:< S-kur Ekur (temple name [TN])

ki...i!gi!2 to love

Notes
E -kur This was the main temple of Enlil in Nippur, by far his most important sanctuary.
It was in the Ekur that the assembly of the high gods would meet, as occasion demanded.
The god An presided over these meetings, but it was the responsibility of Enlil to carry out
the decisions. One of the reasons meetings were held was to select the rulers of Meso
potamia.

ki.. .ag The verb meaning "to love" is written both

and m. Most
Sumerologists believe that both writings represent /ki ... aga/. The first writing is to be
understood as: ki ...i!2 ' and the second as: ki ...M-gA. This means that the sign has
two values: i!g and i!gi!2 . The problem of the "overhanging vowel" in such cases will be
discussed in Lesson 1 1 .
It is not clear what the two elements of this compound verb mean. The most common
meaning ofi!gi!2 is "to measure" (the Akkadian equivalent is madadu), but it can also mean
"to mete out". The most common meaning of ki is "earth". It is hard to say how an
expression like "to measure the earth" could come to mean "to love". Either the word ki,
"earth", had some other meaning, not known to us, or else the element ki in kL .Mi!2 is an
entirely different root ki (that is, a homonym).
The most common Akkadian equivalent ofki...i!gi!2 is ramu, "to love".

95
96 Lesson S

Text 8

Ef;>- . FlI
l fJ =-

--... * rEI
5 4f


I!ffir=> ,.-JJ
" ID) ;lmEJ

==1
d
JO H >-
 Lesson 8 97

Transliteration Transcription Translation

1 : dEn-lil Enlil For Enlil,
2: lugal-kur-kur-ra luga1.kur.kur.a(k) the king of the lands,
3: lugal-i!:-ni luga1.ani.(r) his king -
4: Ur_dNammu Urnammu Ur-N ammu,
5 : nitab-kalag-gi!: nitabkalaga the mighty man,
I-Unm
6: j!g.f!:...
I ki-ma luga1.Urim.a(k) the king of Ur,
S
7: lugru-Ki-en-gi-Ki-uri-ke4 luga1.Kiengi.Kiuri.k.e the king of Sumer and Akkad -
8: :e-kur Ekur the Ekur,
9: -ki-g-g-ni e.ki.aga.a.ani.0 his beloved temple -
10: mu-na-du mu.na.(n.)du.0 he built.
Commentary
9. In Text 2, the adjective kalag-gi!: occurred, derived from the root Ikalagl with the
nominalizer .a. In Text 8, the same formation occurs, but from a compound verb. ki...i!:g
is "to love"; the adjective "beloved" is [ki.aga] .a. The scope of the nominalizer here
involves both the nominal and verbal elements of the compound verb; therefore, the
nominalizer is given in transcription, unlike the case with "simple" adjectives such as
Ikalaga/.
As discussed in Lesson 1 , the third person possessive-suffix after vowels appears
both as i!:-ni and as ni. Perhaps, then, this form should be understood as: ki.aga.a.ni,
instead of: ki.aga.a.ani.

Discussion: structure
The structure of this text is:

[Enlil, luga1.kur.kur.a(k), luga1.ani]. (r) benefactive

[Urnamrnu, nitab.kalaga, luga1.Urim.a(k), agent
luga1.Kiengi.Kiuri.k] .e
[Ekur, e.ki.aga.a.ani] .0 patient
mu.na.(n.)du.0 verb

- Door sockets

Many texts inscribed on door sockets have been preserved; for a photograph of such a
stone, see Text 17a below. Such stones were partially underground, and were used to hold
a door. In Woolley's words,
The Sumerian door consisted of a wooden leaf fixed to a pole rather higher
than itself; the projecting top end was held by and revolved in a metal ring
attached to the lintel, the lower end was shod with metal and went down
through a hole in the pavement to rest and turn on the hinge-stone. This was
98 Lesson S

a boulder of (imported) hard stone, limestone or diorite, in which a cup

shaped hollow had been cut to take the pole-shoe, and generally one part of it
had been smoothed and inscribed with the name of the king who dedicated
the building and of the god in whose honour he built it. ... Imported stones
were valuable and an old stone would often be taken away and re-used for
some building other than that for which it had been intended, so that the old
inscription no longer applies ( 1 982: 1 60- 1 6 1 ) .
In some cases, door sockets were re-used for reasons of piety; in other cases, i t was
purely for reasons of economy.
As Woolley's description implies, the building inscription inscribed on the door socket
would not normally be exposed to view. Royal inscriptions in general were not meant to be
seen by contemporaries of the builder, but rather by future rebuilders.
In Hallo's scheme, door sockets are a subdivision of building inscriptions, which are
themselves one of the major divisions of royal inscriptions.
 Lesson S 99

Text 8 a
supplementary

Another brick.
 Lesson 9

This text is another brick.

Sign-list and vocabulary

Unug (Unu) Uruk (ON)

OO <Hf
E-temen-ni -guru3 Etemenniguru (TN)

13== dumu son

xrJ:=f sag head

en lord

g4 to return, to restore

bi, be

Notes

Unug One of the most important cities in southern Mesopotamia; it often played a role in
political history. Before becoming king, Ur-Nammu had been military governor of Ur
(akkana-Urim5kCma) under the control of Utu-Hengal in Uruk.
The etymology of the name is unknown; this is discussed below. The pronunciation
of the name as IUnugl is known from syllabic writings. However, Semitic spellings show
Ir/: Biblical Erech, modern-day Arabic Warka, etc.; the lexical lists also give the Akkadian
equivalent with Ir/ : U-ru-uk. It is not known why the Sumerian form shows 1nl while
the Semitic forms show Ir/.

E -temen-ni-guru3 This is the name of Ur-Nammu's ziggurat at Ur, illustrated in Lesson

1 . The etymology is uncertain. It is sometimes transliterated as E-temen-ni -ila2'

sag This is literally "head". However, it not infrequently forms the second element of
noun-noun compounds, where, as in many languages, it can take on metaphorical uses.
For example, ka-, "beer-head", is "top quality beer"; dumu-, "son-head", means
"eldest son".
has several Akkadian equivalents: reu, "head" and several derived meanings;
q.!!qqadu, also "head" and several derived meanings; amelu, "man", lu, "young man", etc.

en This is normally translated "lord", a purely conventional translation. The ruler of Uruk
is always called an en. Its meaning is further discussed below.

101
102 Lesson 9

Text 9
 Lesson 9 103

Transliteration Transcription Translation

1 : dNanna Nanna For Nanna,
2: dumu- dumu.sag the eldest son
3: dEn-lil-hi Enlil.a(k) of Enlil,
4: lugal--ni lugal.ani. (r) his king -
5: Ur_dNammu Urnammu Ur-Nammu,
6: nitab-kalag- nitabkalaga the mighty man,
7: en-UnugkCg en.Unug.a(k) the lord of Uruk,
8: ...! Ski-ma
I !gIDI-Unm lugal. Urim.a(k) the king of Ur,
9: lugal-Ki-en-gi-Ki-uri-ke4 lugal.Kiengi.Kiuri.k.e the king of Sumer and Akkad -
1 0 : E-temen-ni -guru3 Etemenniguru the Etemenniguru,
1 1 : -ki-g-g-ni e.ki.aga.a. ani. 0 his beloved temple -
12: mu-na-du mu.na.(n.)du.0 he built;
1 3 : ki-be mu-na-g4 ki.bi.e mu.na. (n.)gi.0 he restored it to its proper
place.
Commentary
1 3 . ki-be mu-na-g4 means "he restored it to its place", that is, rebuilt the temple at its
original location. This entire expression was borrowed into Akkadian as: ana aSrBu itOr,
"he returned (it) to its place" .
. bi is the inanimate possessive-suffix, "its", referring back to the temple. It is the
inanimate equivalent of the animate possessive-suffix .ani.
The most obvious interpretation of the first two words of this line would be to read ki
bi, interpreting it as the direct object (patient) of the verb g4 : "he returned its place".
However, there is evidence that ki-bi here is not a direct object (patient), because variations
of this formula already include an expressed patient. For example, -ni ki-be mu-na-g4,
"he restored his temple to its place"; bad-bi ki-be4 mu-na-g4 , "he restored its wall to its
place".
One interpretation is to read ki-bi, for ki.bi.(se), that is, a terminative nominal phrase.
The terminative .Se becomes IS! after a vowel, and since Isl is an amissable consonant, it
would not appear in the script. While it is true that the morphology of the terminative case
marker is not entirely clear, one might expect to find at least one case in early texts of the
terminative appearing in this formulaic phrase, but none apparently occur; note that it
regularly occur after a vowel in forms such as nam-til-Ia-ni-se. It is true that in later texts
ki-bi-se occasionally occurs in somewhat similar contexts; however, it is possible that this
is due to influence from the Akkadian equivalent.
The second possibility is to read ki-be, for ki.bi.e. The .e is the marker for a case not
yet seen, the "locative-terminative" case. It is difficult to pin down a specific function for
this case. It shares some of the characteristics of the locative case, and some of the
104 Lesson 9

characteristics of the terminative case. In this particular fixed expression, the force
conveyed by the locative-terrninative is: "He returned the temple to its original place". That
is, its meaning here is close to that of the terminative marked by -e.
The form of the locative-terminative dimensional-prefix is much open to dispute. It is
often not cross-referenced at all in the verbal prefix chain. Other times, it seems to use the
dimensional-prefix corresponding to that of the locative case, that is, .ni. (According to
some Sumerologists, .ni in such contexts is for *n-e; the .e is the dimensional-prefix of the
locative-terminative, and the .n is an optional pronominal-prefix, discussed in Lesson 1 5.)
In this particular case, it is not cross-referenced at all in the verbal chain.
The .0 cross-references the patient. There is no expressed patient in this sentence; the
logical patient is the nominal phrase in line 1 1 . Similarly, the logical antecedent of the
datival dimensional-prefix .na is the datival nominal phrase in lines 1 -4.

Discussion: structure
The structure of this text is:

[Nanna, dumu.sag.Enlil.a(k), lugal.ani] .(r) benefactive

[Urnammu, nitab.kalaga, en.Unug.a(k), agent
lugal.Urim.a(k), lugal.KiengLKiurLk] .e
[Etemenniguru, e.kLaga.a.ani] .0 patient
mu.na.(n.)du.0 verb
[kLbi] .e place
mu.na.(n.)gL0 verb

- Sign formation

Over time, the Sumerian writing system produced a number of new cuneiform signs;
the history of these developments is rarely visible to us. One such process was the addition
of short strokes to an already existing cuneiform sign, thereby modifying its meaning. For
example, the word sag, "head", is represented by a sign which was originally the picture of
the head and upper torso of a man: x:..CY::-f ' The word ka, "mouth", is represented by the
same sign, but with the addition of short strokes over the region of the mouth: lf...:r .
Akkadian scribes referred to these extra strokes as gunO. This is an adaptation of a
Sumerian word gunu, apparently meaning "colored". Modern scholars will sometimes
refer to, e.g., the ka-sign as "SAG+gunG". This convention is sometimes necessary, when
the value of a "gunOfied" sign is not known.

- Functions of .e

The locative-terrninative case is marked in .e, and the agentive-ergative case is marked
in .e. An obvious question which springs to mind is, are the two related? It is probably no
accident that the two case-endings share the same phonological shape, / e/. The parallel has
been made to the English preposition "by". This can express a locative ("by the river"), an
instrumental ("by the hammer"), or an agent ("by the man"). The agentive marker .e may
have developed out of this locative-terminative .e. This .e may have started to lose some of
its functions, which began to be taken over by the locative case in .a and by the terminative
 Lesson 9 105

case in .e. Haayer has discussed this point, in terms of universal tendencies in language:
One of the most characteristic features found in case marking in ergative
languages is that the ergative case is often identical with another case, most
often the genitive or instrumental, sometimes the locative or dative. In
Sumerian, for instance, the ergative case is marked by the postposition -e,
which is identical to the locative-terminative -e, and is in origin a deictic
pronoun ( 1986:80).
Although we understand and translate simple Sumerian sentences such as lugal.e e.0
mu.n.du.0 as "the king built the house", it is usually assumed that at some "Proto
Sumerian" stage, the meaning may have been something like "a house got built, connected
with the king", or "there was a building of a house by the king". That is, to some degree,
pre-historic Sumerian (and historic Sumerian?) should be understood as basically "passive"
in nature. It has been said that in ergative languages, the patient of the sentence is the
"topic" of the sentence (while the agent is the "comment"), but in accusative languages, it is
the subject which is the topic (while the patient is the comment).

- Noun compounds

dumu-sag is literally "son-head". This represents a case of noun-noun compounding.

This is not a productive method of word formation in historic Sumerian, but a few such
cases exist. Several early proper names are noun-noun compounds: dEn-Hl, "lord-wind";
E-kur, "temple-mountain", etc.; these are not genitive formations.
The element nig, used to form concrete (and occasionally abstract) nouns from verbal
roots, is in origin a noun meaning "(some)thing". The original meaning of nam is less sure,
but possibly had a similar origin, or may have meant something like "state-of-being".
Thus, abstract and concrete nouns formed from either Dig or nam originally represented
noun-noun compounding.

- History

The different functions of the en and l!!g al have been much discussed; they varied to
some degree from place to place and from period to period. In Jacobsen's seminal article on
"Early Political Development in Mesopotamia", he stated that in the earlier periods the en
(Akkadian blu) was more of an "administrator" while the l!!g al (Akkadian arru) was a
"warleader":
In the case of the en the political side of the office is clearly secondary to the
cult function. The en's basic responsibility is toward fertility and abundance .
... The "king", l!!g al, in contrast to the en was from the beginning a purely
secular political figure, a "warleader" ( 1 957 :375 n.32).
As for the more "original" meanings of the telms,
The Sumerian term en which is generally translated "lord" denotes basically a
productive manager, someone with magic gifts to make things thrive
( 1 987:20 n.2; cf. p. 277) . ... Under the early political forms ... the king
(l!!gil)l was usually a young man whose task it was to lead the army in war
( 1 987:236 n.4).
106 Lesson 9

- Substrate

The etymology of the name Uruk is unknown. Many of the oldest cities in Sumer
have names which are not apparently Sumerian. Such names go back to the language(s)
spoken by the people(s) living in southern Mesopotamia before the Sumerians arrived; the
name of the city of Ur may be one of these names. Gelb says:
Almost all the Mesopotamian geographical names found in the earliest
Sumerian sources are non-Sumerian and non-Akkadian and must be
assigned to the proto-population of Mesopotamia. This conclusion is true of
the names of rivers and mountains, as well as of cities and countries. Only in
the Pre-Sargonic period do we find the first attestation of Sumerian
geographical names ( 1 962:49).
(A number of these place-names are discussed in Limet 1 97 5b.)
These substrate peoples, about whom virtually nothing is known, were referred to by
Benno Landsberger as "Proto-Euphrateans"; they are also sometimes referred to as "Proto
Tigridians". The words for certain material objects and certain professions in Sumerian go
back to this language, for example, nagar, "carpenter", which has no obvious Sumerian
etymology. Some of these substrate words then passed on to Akkadian, and eventually on
to Aramaic, Hebrew, and Arabic.
Scholars disagree in their views as to how much of Sumerian vocabulary is of
substrate origin. Some are inclined to see a large number of substrate words in Sumerian,
including many place-names and divine-names; other scholars are less convinced.
Jacobsen, for instance, has Sumerian etymologies for several city-names, which other
scholars regard as substrate names. There is rarely enough evidence to decide any
particular case.

- History

Control of Uruk was important to all the Ur HI kings. Hallo says:

Certainly the two cities [Ur and Uruk] had a venerable history of dynastic
and administrative union behind them . ... Ur under Ur-Nammu was heir to a
long history of dynastic and administrative union with both Uruk and
Lagash . ... Nippur is the religious center, Ur the political capital and Uruk,
from all indications, the ancestral home of the dynasty ( 1 966: 1 36- 1 38).

- History

It i s difficult to say how literally the expression "he restored it to its place" should be
understood. The Sumerian phrase (and the corresponding Akkadian phrase) is somewhat
ambiguous; it can mean either "to restore to a former spot" or "to restore to a former state".
Kings of the Neo-Babylonian period specifically claimed that they took care to rebuild
temples exactly on old foundations. Woolley says that
It was customary in Mesopotamia, when rebuilding a temple, to incorporate
the earlier one within the core of the platform upon which its successor was
to be set. This often meant largely dismantling it ( 1 982: 1 09).
 Lessoll 9 107

Nabonidus. for example. has left several inscriplions in Ur, in which he claims to have
restored the ziggurat of Ur-Nammu (the Eternenniguru). He states. in fact. that Ur-Nammu
started the work on the ziggural, but did not finish it; Ur-Nammu's son and suc",essor
Shulgi also worked on the complex, but did not finish it; only he. Nabonidus himself,
completely finished and restored it. The following drawing is WooJley's reconstruction of
Nabonidus' ziggurat; it is instructive to compare i t with Woolley's reconstruction of Ur
Nammu's ziggurat, given in Lesson 1 . Ur-Nammu's ziggurat itself was built over an earlier
temple, which itself was built over an even earlier temple.

THE ZICCURAT OF NABONIDUS RESTORED

....,. PO 8.C

_ . - .. ...
108 Lesson 9

As another example of temple-rebuilding, the Ishtar Temple of Assur was in existence

some two thousand years, and was frequently rebuilt. Richard Ellis says that "It was
always in about the same place, though sometimes the new version would be placed to one
side of the earlier ruins" (1968 : 1 2). The temple of Inanna at Nippur had an even longer
history. It was built and rebuilt from at least as early as Early Dynastic I to late Parthian
times - that is, about 2700 BC to 150 AD. There were at least eleven major building levels.
The temple varied in size from period to period, usually getting bigger. The new
sanctuaries were normally built over the previous ones.
The principles behind the orientation of Mesopotamian temples are not at all clear,
especially in the older periods; some of the evidence seems contradictory. Nor are the
means by which the Mesopotamians determined the orientation known. Ganther Martiny,
writing in 1940, says that
Astronomical orientation is ... especially noticeable in the case of late
temples. The direction of orientation should probably be understood as the
direction in which the god's statue faced. ... In Neo-Babylonian times
orientation based on individual stars assigned to specific deities came into
vogue (1940:92).
Sally Dunham, however, writing almost fifty years later, is less sanguine:
Very little is known about how the ancient Mesopotamians oriented and
measured off the ground plans and precincts of their temples, although we do
know such measuring was important enough to be mentioned in their royal
inscriptions and religious texts. ... Still today nothing is known about if and
how the ancient Mesopotamians used astronomy to orient their temples
( 1986:39, and n.37).
Martiny thinks that the "Gimilsin (Le., Shu-Sin) Temple" in Eshnunna was oriented
toward the city of Ur:
Exactly along the projected axis of the Gimilsin Temple in the direction in
which the god's statue faced, at a distance of about 300 km. toward the
southeast, lies Ur, the residence of Gimilsin. Is it possible that the deified
ruler, in whose honor the temple in Eshnunna was to be built during his
lifetime, had demanded orientation of the temple toward Ur? ... The Gimilsin
Temple confronts us with what appears to be a case of geographical
orientation toward the capital of the overlord (1940:95-96).
 Lesson 1 0
This is a text of Ur-Nammu's son and successor, Shulgi, who ruled from 2094 to
2047 BC. It is inscribed on a weight, in the shape of a sleeping duck.

S i gn list and vocabulary

-

?:mrr Sul-gi S hulgi (PN)

* 1l an-ub corner (?)
R da side
limmu2 four

Notes

S ul-gi In older transliterations the name was read as Dun-gi. It is almost always read Sul
gi nowadays, although there is really very little evidence to permit a decision one way or the
other. It is usually interpreted as "noble (gir) young man (ul)".
The gi-element had an Ir/-Auslaut, although the standard sign-lists do not record any
value in Igir/. Because of this Ir/-Auslaut, he is occasionally referred to as "Shulgir".
This gi-element may be the same gi-element seen in the GN Ki-en-gi.

an-ub The analysis is not clear. Because this word is occasionally written without the an
sign, some think that the an-sign is the divine determinative, and so it is sometimes
transliterated as: dub. Sollberger, for example, explains the word as: "part of the world
(as an emanation of the divine, hence the (divine) classifier)". However, these omissions of
the an-sign occur only in relatively late texts. For simplicity sake, it is transliterated here as:
an-ub, and transcribed as: anub.

da The cuneiform sign is a picture of the head-upper-shoulder-arm. Its meaning was

extended to mean "side". It is equated in lexical texts with the Akkadian idu, glossed by the
CAD as: " 1 . arm, 2. side, edge, border, ... 7. strength".

limmu2 Numbers can be expressed in two ways in Sumerian: either by a numeral, or by

spelling out the number, using a mixture of logographic and syllabic signs. The word for
"four" in Sumerian was pronounced Ilimmul (or perhaps llimu/, or llima/). This number
was usually expressed by its numeral: it . This numeral can be transliterated as limmu.
However, when Sumerian numerals are used strictly for counting, they are usually
transliterated by Arabic numerals, e.g., "4". In certain contexts, when not used strictly for
counting, "four" is expressed by limmu2 ' as in Text 10.

1 09
1 10 Lesson 10

Some read the sign as limu2 instead of limmu2 . In older transliterations, it is fre
quently transliterated as tabtab. This is still preferred by some modern-day Sumerologists.

Text 1 0

t7
m II
* fHlli
tf f fIrt>
t=C] 5

< 8
*-
rt:r

Transliteration Transcription Translation

1 : dNin-gal Ningal For Ningal,

2: nin--ni nin.ani.(r) his lady -
3 : dS ul-gi S ulgi Shulgi,
4 : nitab-kalag-g nitabkalaga the mighty man,
5 : J!g
I Ski-ma
I- Unm lugal. Urim.a(k) the king of Ur,
6: lugal-an-ub-da-limmu2 -ba lugal.anub.da.(k) limmu.bi.a(k) the king of the four
quarters.
 Lesson 10 HI

Commentary
1 -2. The text begins with a benefactive to a goddess, even though there is no votive object.
The use of the benefactive is due to the formulaic nature of these inscriptions, coupled with
the idea that all official activity, including the regulation of weights and measures, was
conducted ad maiorem deorum gloriam.
3. At some point in the middle of his reign, Shulgi's name begins to appear with the
determinative normally reserved for divine names, *" . This (and other) evidence indicates
that Shulgi was deified, both during and after his life-time; this is discussed further below.
6. There are two genitive constructions in this line, one of which has not been seen
previously: the "anticipatory genitive". Up to this point, to express "the temple of the
king" in Sumerian, one would expect e.lugal.a(k), written -h!gal-la. However, Sumerian
has an alternative genitive construction, which puts the noun with the genitive marker first,
followed by the second noun with a possessive-suffix. A literal translation of this
construction would be: "of the king, his temple". This would be: lugal.a(k) e.ani, written
either lugal-la -i!-ni or lugill-la -ni.
This particular expression loosely translates as "the four quarters". A literal translation
would be: "of the corner-and-side, its four": anub.da.(k) limmu.bi. As in the expression
"king of Sumer and Akkad", there is no conjunction between "corner" and "side". Since
these terms are inanimate, the form .bi is used, not .ani.
However, this entire expression is itself the second element of a regular genitive
construction: "king (of the four quarters)". The first element in this genitive phrase is
h!gal. The second element of this genitive phrase is the entire phrase: anub.da.(k)
limmu.bi. The second element is then followed by the genitive marker .ak. The lil of the
possessive-suffix .bi contracts into the tal of .ak, producing Ibak/, and the Ik/, as is the
normal practice, is not written. Thus, a literal translation of this entire expression would be:
"king of [of the corner-and-side, its four] ": [lugal]. [anub.da.(k) limmu.bi] .a(k), producing
"king of the four quarters".
The anticipatory genitive tends to occur in certain fixed expressions (such as in line 6).
In theory, it can be used anywhere a regular genitive could be used, but in practice it is less
common. Since the expression "king of the four quarters" is quite frequent, it is not a
problem to recognize it in context. However, non-idiomatic uses of the anticipatory
genitive can be quite difficult to recognize. The two clues for its presence are: an otherwise
unexplained la/-vowel, followed a little later by an otherwise unexpected possessive
suffix. Several instances of the anticipatory genitive occur in the following texts.

Discussion: numbers
One of the lexical texts found at Ebla is a small tablet giving the names of the Sumerian
numerals from one to ten, spelled more-or-less syllabically. This tablet (TM.75.G.2 198)
,
was apparently some kind of school or practice text. For "four", the tablet says: li-mu,
presumably for Ilimmu/.

- Typology

It is more common for S-O- V languages to have a genitive construction of the type
112 Lcssoll lO

possessorpossessed then of the type possessed-possessor. The genitive in Turkish, for

example, (an S-O-V. agglutinative language) is of the type possessor-possessed. Sumeri3n
is rather unusual in that the most common pattern is possessed-possessor ("the palace of
the king"), The anticipatory genitive. however, is of the type possessor-possessed ("of the
king, his palace"), and it is possible that it represents the older construction, which perhaps
was in the process of becoming limited to certain stock expressions.

-Weights

Text 10 is a standard inscription, as was Text 7a1b. However, it is inscribed on a

weight. This particular weight does not bear any indication of its value. Text 21a, on the
other hand, is a weight inscribed with its value: mana glna, that is, "5 standard minas".
Weight inscriptions, such as this one, were typically carved of stone in the shape of a
sleeping duck. It is not clear why such a shape was used. The following illustration is of a
duckweight from the NeoAssyrian period.

Although stone was the usual material used for making weights, in Assyria bronze
weights were sometimes used instead.
 Lesson 10 1 13

- History

The first Mesopotamian ruler to use the divine determinative before his name was
Naram-Sin, the fourth ruler of the Dynasty of Akkad, who ruled approximately 2254-22 1 8
BC. Gadd says that "no doubt the vast accession of power and width of sway won by
such a mighty figure as Naram-Sin helped to make him appear superhuman" ( 1 97 1 :619).
The determinative was used by all the Ur III rulers except Ur-Nammu. It was used only
sporadically by following rulers.
Occasionally, epithets in the royal inscriptions use the word "god". In Text 1 7a, for
example, Amar-Sin refers to himself as digIT-zid dUtu-kalam-ma-na, "the effective god, the
sun-god of his land".
There is also a certain amount of literary material which indicates that the Ur III kings
were considered, or considered themselves, "deified". However, it is not really known
what this means. The use of the English word "deified" is rather facile; it is very difficult to
say what this meant to the Ur III rulers or to their subjects. (Gadd says that "vainglory and
popular superstition supported [this policy]" [197 1 :6 19].) However, there is a certain
amount of evidence from royal tombs of the Ur III period to indicate that offerings were
made to the dead Ur III kings, implying that they were worshipped as gods after their
death.
Moorey says that
The most common evidence for the worship of the deceased kings of the Ur
III Dynasty is provided by economic documents describing deliveries to a
place called ki-fl-Mg, where liquid offerings to the dead were libated. No
thing specific is known of these mortuary shrines (1984: 17).
This topic of the deified king has been discussed by Jacobsen:
The deified king is not a "god" generally; he has the specific relation to the
country that a personal god has to his ward . ... The king, as leader of the
country and originator of policy, is the "personal god" of his realm. The
deification of rulers in Mesopotamia is accordingly to be understood not in
terms of the qualitative contrast human:divine, mortal:immortal, etc. , but in
terms of function of the king, he is the "genius" of the country ( 1 957: 395
n . 1 08).
Although the divine determinative is never used before Ur-Nammu's name in any of
his royal inscriptions, it is so used in the Prologue to the Law Code usually ascribed to him.
But this is probably the work of Shulgi, and in any case the Prologue is a rather late copy,
dating from the Old Babylonian period. Also in the Prologue, he is referred to as the son of
the goddess Ninsun; he is elsewhere referred to in the same way.

-Titulature

It was mentioned above that Naram-Sin of Akkad was the first Mesopotamian ruler to
use the divine determinative. He was also the first to use the title "king of the four
quarters", in both an Akkadian form and a Sumerian one. The title was not apparently used
by other Akkadian kings. It was used once by a Gutian king, and once by Utu-Hengal of
Uruk, who was overthrown by Ur-Nammu. Ur-Nammu himself did not use the title,
1 14 Lesson 10

presumably because of the limited size of his realm; note also that he did not use the divine
determinative. The title was used by all the other Ur III kings, and afterwards by various
later kings, in an Akkadian or Sumerian form.

- History

Ur-Nammu was killed on the battlefield, but no specifics of his death are known; the
literary work entitled "The Death of Ur-Nammu" is terse and unsure at this spot. Woolley
believed that Ur-Nammu and the other Ur III rulers (except the last) were buried at Ur, in a
building complex he referred to as the "Mausolea" of the Ur III rulers. Moorey has
recently questioned this:
The balance of available information, archaeological and textual ... suggests
that if the kings (and queen-mothers) of the IIIrd Dynasty of Ur were buried
in that city it was not in Woolley's "Mausolea" . ... Ur is not the only potential
site for these graves, for they might have been in a palace at Uruk, home of
the dynasty, or, less probably, even perhaps at Nippur ( 1 984: 1 8).
It was under the rule of Ur-Nammu's son and successor Shulgi that the Ur III empire
reached its greatest extent; Piotr Steinkeller calls him "the true builder of the Ur III state"
( 1 987b:20). There was a great deal of royal building; there was a reform in the calendar,
and much bureaucratic reorganization; Gadd says that "visibly under the impulse of the
king himself a most meticulous system of bookkeeping was instituted" ( 1 97 1 :617). Daniel
Snell also points to "a general economic stability during the middle years of the Ur III state,
a stability that seems a likely corollary of the middle kings' largely successful attempts to
maintain the empire their predecessors had bequeathed them" ( 1 982: 1 9 1).
Steinkeller ( 1987b:20-21 ) lists the following among the reforms of Shulgi:
1) the deification of S hulgi
2) the creation of a standing army
3) the reorganization of the system of temple households
4) the creation of a unified administrative system for southern and northern
Babylonia
5) the introduction of the bala taxation system, coupled with the creation of
a chain of redistribution centers ... which served to collect, to process,
and to distribute the state revenues
6) the creation of an enormous bureaucratic apparatus, as well as of a
system of scribal schools that provided highly uniform scribal and ad
ministrative training for the prospective members ofthe bureaucracy
7) the radical reform of the writing system
8) the introduction of new accounting and recording procedures and of new
types of archival records
9) the reorganization of the system of weights and measures
1 0) the introduction of a new calendar, the so-called Reichskalender, which
became the official caldendar throughout the Ur III state.
It was about half-way through Shulgi's rule when he began to conduct many military
raids. A number of these were directed towards the East, modern-day Iran. The details of
 Lesson 10 1 15

these campaigns are rarely known to us, and in fact it is surprising how little historical
information we have about Shulgi, especially considering the fact that he ruled for almost
half a century.
Shulgi was also the subject of some thirty hymns, preserved to varying degrees. In
"Hymn B", he boasts: "I learned the art of the scribe from the tablets of Sumer and
Akkad"; he also refers to himself as "the scribe of Nisaba", the goddess of wisdom and
writing.
1 16 Lesson 10

Text l Oa
supplementary

On a brick from the centre of the R u ins

Notes

This autograph is taken from Volume 1 of Sir Henry Rawlinson's The Cuneiform
Inscriptions of Western Asia, published in 1 8 6 1 . This volume was one of the earliest
collections of cuneiform inscriptions published. At that time, very little was known about
Sumerian. In fact, it seems that it was not until the year 1 869 that the word "Sumerian"
was attached to this language, by Jules Oppert.
It is very difficult to correctly copy texts written in a language which one does not
understand. This means that such autographs are sometimes slightly "off '. In this
particular autograph, some of the signs seem to differ from those in the other texts.

7. en-UnugkC !@: Funny looking Unug-sign!

The shape of the ki-determinative in this line is quite different from the shape of the ki
sign in lines 9 and 1 1 . It is hard to say how much of this variation is due to the original, or
how much is due to Rawlinson.

8. There is no expected -ma at the end. This is most unusual, and one suspects that this
is an error of Rawlinson, rather than an error of the original scribe.
 Lesson 1 1
This text was inscribed on a headdress or wig of diorite. D.J. Wiseman has pointed
out that this headdress was "scored on the underside to fit the rounded head of a statue".
He concludes that it was "intended to be fitted to an actual statue rather than be used
independently as an ex-voto object" (1960: 168 n.25).

Sign-list and vocabulary

t::tt Lamar Lamar (DN, fern)

w=r ;@ Nin-gir-su Ningirsu (DN, masc)

Nane Nanshe (DN, fern)

x8 Ell Ba- Baba (DN, fern)

* (j- at. Bfj Tt *" dBa--nin-am Babaninam (PN)

IH *" Bfj Ur- dNin-gir-su Ur-Ningirsu (PN)

t[Jl- zabar bronze

t[J l- B zabar-dabS (kind of official)

<> bi-li attraction; headdress, wig

munus woman

8 dabs to hold

dim to fashion, form

t\rT
EM y, ba6

n* am

Notes

Lamar The reading of the name in the first line is unclear, since the text is partially broken.
There is a commonly attested goddess whose name is usually written with the kal-sign (Le.,
the kalag-sign); it is variously read by Sumerologists as Lama2 (or Lama), Lamma2 (or
Lamma), Lamar, or Lammar. Since this word was apparently borrowed into Akkadian as

1 17
1 18 Lesson 1 1

lamassu, there was probably some kind of Ir/-Auslaut; Lieberman reconstructs the original
form as: ILamat/. (The word is further discussed below.)
Virtually everyone who has studied this text has read the first line as dLamar (re
gardless of the precise transliteration). However, there seems to be present a stroke of
another kal-sign, immediately after the divine determinative: d [Kall-kal. This may be a
different writing for the same goddess Lamar, or it may represent an altogether different
deity (a god dKal-Kal is elsewhere attested, but seems to be masculine). For simplicity
sake, and since only a single stroke remains of the problematic sign, the line will be
transliterated as dLamar.

Nin-gir-su Etymologically, "Lord of Girsu". nin is used here in the sense of en, "lord",
as in Text 1 .
Girsu was the sacred quarter of the city and state of Lagash (further discussed in
Lesson 22); it actually lay some distance outside the city proper. Ningirsu was the tutelary
divinity of the state of Lagash. His most famous temple was the E-ninnu (of uncertain
meaning, "House 50" ?).
Ningirsu seems to have been the local name for the god elsewhere worshipped as Nin
urta, a god originally of agriculture and storms, but also of war. The two were probably
independent deities who were very early identified with each other.

Nane This name is read by some Sumerologists as Nazi. She was a daughter of Enki,
and the goddess of Lagash. She was consulted for the interpretation of dreams. When
Gudea, the ruler of Lagash, had an odd dream in which a mysterious figure appeared, it
was Nanshe that he turned to for the explanation of the dream.
The cuneiform sign representing her name is the sign for the city of Sirara (one of the
places where she was especially worshipped), with an inscribed -sign. means
"fish"; the sign is in origin the picture of a fish. This and other evidence indicates that
Nanshe may originally have been some kind of fish-goddess.

Ba-ba6 The reading of the last sign is uncertain. The name is variously transliterated as:
Ba-y, Ba-, Ba-bU l l (and Ba-bu l 2 ) ' and Ba-w,\. She was the wife of Ningirsu, and
hence the city-goddess of Lagash. At times, Inanna herself is referred to as "Ba-".

zabar In older transliterations, each of the three individuals signs forming this compound
logogram is separately transliterated: ud-ka-bar. The etymology and writing are discussed
below.

zabar-dabS The etymology is discussed below. The function of this official is not too
clear. Jacobsen refers to him as "the official in charge of the bronze (table-wares, cups,
knives etc. of a large establishment, and possibly of the bronze weapons as well)"
( 1 957 :382 n.55).
The word was borrowed into Akkadian as zabardabbu. The CAD simply translates
the Akkadian term as: "an official". After a long discussion, it concludes with the remark
that this official was "(possibly), originally the weapon carrier of the king".
 Lesson 1 1 1 19

bi-li The basic meaning of bi-g appears to be something like "chann" or "attraction". (The
Akkadian equivalent, kuzbu. is glossed by the CAD as: "luxuriousness, abundance,
attractiveness, charm, sexual vigor".) It also has the derived meaning "headdress" or
"wig". The gudug-priests discussed in Lesson 1 9 are attested as wearing a bi-g.
munus Also transliterated as mi . Particularly in older works, it is transliterated as sal.
dabS In older transliterations, dib (and, incorrectly, dib4).
dim While du is used for the (re)construction of more solid objects, such as palaces,
temples, etc., dim is nonnally used of smaller, hand-made objects.
The Akkadian equivalents for du and dim are not neat. dim is nonnally equated with
'
banu, but du is equated both with banu and epeu.
Am This sign is composed of two elements: Tt which nonnally has a syllabic reading ,
'

and * which nonnally has a syllabic reading an. It is not clear how these two signs
'

came to represent (together) the value I amI.

120 Lesson 1 1

Text 1 1

4ClFr ET
rrIJ1I
r1! .: ;:'
. . .

*:'
O
+-f r4 k'=l
 Lesson 1 1 121

Transliteration Transcription Translation

1 : dLamar Lamar For Lamar,
2: nin-.f!-ni nin.anL(r) his lady -
3: nam-til nam.til for the sake of the life
4: dS ul-gi S ulgi of Shulgi,
5: nitab-kalag-g! nitabkalaga the mighty man,
6: lugill-Urimski-ma-ka-e lugal.Urim.ak.a(k).e the king of Ur
7: d--1!
B b 6" nm-am
" Babaninam Babaninam,
8: zabar-dabS zabardab the zabardab
U d---1!l
9 : J- -,
N' -grr-su Umingirsu of Ur-Ningirsu,
10: en-ki-.f!_dNane-ka-ke4 en.kLaga.a.Nane.k.ak.e the beloved lord of Nanshe -
1 1 : bi-li-nam-munus-ka-ni biILnam.munus.(a)k.anL0 her "beauty of womanhood" -
12: mu-na-dim mu.na.(n.)dim.0 he fashioned.
Commentary
3-6. The essence of this line is: "for the life of Shulgi". This would be expressed by:
[nam.til. S ulgi.k] .e. However, this nominal phrase is complicated by the presence of two

[
appositives. The first is the noun-adjective combination, nitab kalag-gf!.. The second is the
genitive phrase "king of Ur". By itself, this last genitive phrase would be expressed by:

J
lugal.Urim.ak. This nominal phrase may be diagrammed as:
[nam.til] S UIgi .ak.e
nitabkalaga
lugal.Urim.ak
This results in a succession of two genitive markers, followed by the marker for the
tenninative case: .ak.ak.M. In the script, this is reflected as: ... Urimskt ma-ka-e.
Since the Ikl of the first genitive marker .ak is followed by a vowel, it is pronounced
and written in the script. The Ikl of the second genitive marker .ak, however, is syllable
final before a consonant. In such cases, the Ikl does not show up in the script. The
problem of the amissable word-final consonants was discussed earlier. To some degree,
that same problem is present when these consonants are syllable-final, not just word-final.
That it, it is not sure if these consonants were pronounced or not: it is not known if this is
an orthographic or phonological problem. The sequence .ak.ak.e, for instance, is common,
but the Ikl of the second genitive marker does not appear to ever be written in any of these
Occurrences. In general, it seems that the Ikl of the genitive marker does not appear in the
script, when it is in syllable-final position followed by a consonant. It is transcribed here
within parentheses.
The ka-sign includes the Ikl of the first genitive marker, and the lal of the second
genitive marker. This is yet another instance where the script does not follow the
morphology, but rather approaches the syllabic structure of the spoken language. This
122 Lesson 1 1

sequence may well have been syllabified something like /u-ri-ma-ka(k)-se/.

dBa-Ba nin-am is a personal name, "(the goddess) Baba is a lady", or perhaps "Baba,
o
is queen". The divine determinative "goes with" the name of the goddess Ba-ba6 ' not with
the personal name itself.
Sumerian has two ways to express the copula. The first way, which is in fact
relatively uncommon, is to inflect the verbal root meaning "to be" (me); it thus behaves like
a regular verb. This is sometimes called the "independen t copula". The second way, which
is much more common, is to use a reduced form of this root as a suffix, instead of as an
independent verb. This is called the "enclitic copula". For the third person, this consists of
.am (usually written n* am) suffixed to the second element of an equational sentence.
Thus, "Nanna is king" is: Nanna.lugal.am, written: dNanna-Iugal-am; "Baba is queen" is:
Baba.nin.am, written, dBa-ba6-nin-am. (In pre-Ur III texts, the enclitic copula is regularly
written *' read as am6 ' )
To judge from previous writings of morphemes beginning with a vowel, one might
have expected a writing something like dBa-lli!.6 -nin-nam, or d Ba- -nin-na-am. How
ever, this is a case where the script is morpheme-bound; the one sign am regularly
expresses the morpheme .am, and there is no graphic reduplication of the preceding
consonant. (Nor does the writing reflect the presumed syllabic structure.)
In general, names of men in Sumerian are construed with the names of gods, and
names of women are construed with the names of goddesses. However, there are
exceptions to this rule ("Ur-Nammu", for example, is construed with the name of the
goddess Nammu), and there seems to be no evidence that the zabardab-official was ever a
woman. Thus, it cannot be determined whether Babaninam was a man or woman.
8. The word zabar-dab S is composed of two elements, zabar, "bronze", and dab S ' a verbal
root meaning "to grasp". dabS here is probably an active participle. In general, verbal roots
in Sumerian have two participles: an active participle in .0, and a passive participle in .a
(the same nominalizer seen previously). The use of these rather conventional terms is not
without problem, but in general the active participle denotes the doer of the action. Thus,
dab S is "the one who grasps", "he who grasps". The passive participle denotes the result of
the action, or the one acted upon. From sar, "to write", the passive participle sar-ra means
"something written".
Some Sumerologists refer to adjectives in .0 (such as man) as active participles, and
adjectives in .a (such as kalag-gf!.) as passive participles. It is not sure if this is a valid use
of these terms; part of the problem is the difficulty in defining the different categories of
root in Sumerian: verbal, nominal, etc.
As do participles in English, the participle in Sumerian can take a direct object. In this
case, zabar is the direct object of dabs ' Thus, an etymological translation of zabar-dabS
would be: "the one who grasps the bronze", or "he who grasps the bronze" (the object
precedes the participle, just as the direct object (patient) precedes a verb). However, this
particular phrase may have been felt as one unit, since it was borrowed into Akkadian as
one word, zabardabbu.
9. The PN Ur-Ningirsu means "the man of Ningirsu". The DN Ningirsu itself means
"lord of Girsu". Therefore, the name might be understood as: [Ur] .[Nin.girsu.k].ak. The
 Lesson 1 1 123

first genitive marker is for [Nin] . [Girsu] .k; the second is for [Ur] . [Ningirsu.k] .ak.
However, as discussed in Lesson 1 concerning the name Ur-Nammu, there is some
indication that the genitive marker in PNs was lost, and so it is not indicated in
transcription.
Lines 7-10 form the ergative nominal phrase, expressing the agent of the transitive
verb. The ergative case-marker .e appears at the end of line 10. The nominal phrase is
complicated by the presence of all the appositives. Line 7 is a PN, Babaninam. Lines 8-9
ru;e an appositive, describing Babaninam as the "zabardab of Ur-Ningirsu". This would be
expressed as: [zabardab]. [Urningirsu] .k, but Ur-Ningirsu himself is described as "the
beloved lord of Nanshe", an,.. appositive. "Beloved lord of Nanshe" is: [en.ki.aga.a] .
[Nane].k, a more complicated genitive phrase than any seen up to now: ki.aga.a modifies
en, and the combination is itself the first element of the genitive phrase. Thus, the nominal
chain describing Ur-Ningirsu is: Urningirsu, en.ki.aga.a.Nane.k. All of this is the second
element of a genitive phrase, with zabardab being the first element: [zabardab] . [Urningirsu,
en.ki.aga.a.Nane.k] .ak, all of which is an appositive to Babaninam. This may be dia-

]
gr

1;:: b
am
rUrningirsu ]
.e

l
amm .ak
Len.ki.aga.a.Nane.k
The writing of the end of the nominal phrase is as expected: ... dNane-ka-e4.
1 1 . munus is "woman", nam-munus is an abstract, "womanhood".
This line must mean something like "her beauty of womanhood", that is, "her woman's
beauty". If so, then it is a genitive phrase, followed by a possessive-suffix: [bili.nam.
munus.ak] .ani.
The basic rule for the genitive marker as presented up to now has been: /ak/ after
consonants, /k/ after vowels. Here, however, there occurs: nam-munus-ka-ni; the vowel
/a/ of the genitive marker /ak/ does not seem to appear in the writing.
Such writings - where the /a/ of the genitive marker does not appear after a consonant
- are not uncommon; the next example is in a formulaic phrase appearing in Text 13. It is
difficult to say whether such writings tell us something about Sumerian orthography, or
Sumerian phonology, or Sumerian morphology; several interpretations are possible. In
order to make the written form fit our understanding of the grammar, one school of thought
would read the first sign as munusa, instead of munus. This would produce munusa-ka-ni,
accurately reflecting munus.ak.ani. Parallel phenomena occur outside of the Ur III corpus.
For example, in the inscriptions of Gudea - inscriptions highly localized to one time and
place - "his king" is expressed by both lugal-ni and lugal-.!!-ni. In order to make lugal-ni fit
more accurately our understanding of Sumerian, some scholars would read the two signs as
lugala-ni, and not lugal-ni.
This school of thought was particularly adumbrated by Adam Falkenstein, who saw
similar phenomena elsewhere in Sumerian grammar. He coined the term "Uberh!tngende
Vokale" (in English, "overhanging" or "overlapping" vowels) to describe just such
writings. This school would thus see such writings as an orthographic problem. (The
standard sign-lists, however do not seem to recognize a reading *munusa for the sign in
124 Lesson 1 1

question.)
A different view sees this as a morphological (or phonological) problem. Mamoru
Yoshikawa feels that the genitive marker was sometimes Ikl after a consonant, not only
after a vowel. However, he cannot posit rules for the distribution of lakl and Ikl after
consonants. Yoshikawa also believes that there are cases where the genitive marker is
I ak/, not Ik/, after a vowel.
This is not a problem which can be solved here. But it must always be kept in mind
that the writing system of Sumerian never accurately reflected the spoken language. It is
possible that although a scribe spoke /lugalanil, he was perfectly happy to write l!!gal-ni,
because with just these two cuneiform characters, he knew what to read. Why bother to
write an i!-sign if the context makes the presence of a spoken lal obvious?
This entire line is the direct object (patient) of the verb dim, and therefore is in the
absolute case.

Discussion: structure
It is difficult to see the basic structure of this text, because of the presence of so many
appositional phrases. Its substance, however, is: "Babaninam fashioned a wig for Lamar,
for the sake of the long life of Shulgi":

[Lamar, nin.ani] .(r) benefactive

[nam.til. S ulgi, nitab.kalaga, purpose
lugal.Urim.ak.a(k)].e
[Babaninam, zabardab.Umingirsu, agent
en.ki.aga.a.Nane.k.ak] .e
[bili.nam.mun us. (a)k.ani] . 0 patient
mu.na.(n.)dim.0 verb

- Writing system

The innocuous-looking word zabar illustrates some of the intricacies of the Sumerian
writing system. There are no metals native to Sumer; rather, all had to be imported. Thus,
zabar i s not a native Sumerian word; it was borrowed from some unknown language.
Hallo says "in general, it may be supposed that the basic metal names are non-Sumerian
'Kulturworter' or 'Wanderworter' which were adopted together with their referents"
(1963: 140). In Akkadian, the word for bronze is siparru. Akkadian may have borrowed
this word independently from the same language that Sumerian borrowed it from, or, much
more likely, borrowed it directly from Sumerian. In either case, the form siparru is a little
odd; it would seem to derive from */sipar/, not Izabar/.
One way to solve this discrepancy is to assume that in earlier Sumerian the word for
"bronze" was, in fact, Isipar/, and that Izabarl represents an inner-Sumerian development.
The change of Ipl > Ibl is not surprising; voicing of inter-vocalic voiceless consonants
happens in many languages. The change of initial Isl > IzI is less easily explained, but
there are other parallels to this change in Sumerian. The difference in vocalization between
the two forms is more interesting. As will be discussed in Lesson 20, there is a fair amount
 Lesson 1 1 1 25

of evidence to show that Sumerian has undergone a rather wide process of vocalic
assimilation. In words originally containing two vowels of differing quality, one vowel has
assimilated to the quality of the other. In our particular case, an original */i-al has become
la-al.
Thus, Izabarl can be derived from Isipar/, using sound-changes which are elsewhere
attested in Sumerian. (Unfortunately, not enough is known to date these sound changes.)
This then would represent a case where the Akkadian word has actually preserved a more
archaic form of the word than has Sumerian. Presumably Akkadian borrowed it from
Sumerian before these changes took place. Similar instances will be discussed in Lesson
20.
The pronunciation of this word in historic Sumerian as Izabar/, and in Akkadian as
Isiparru/, is known from lexical lists, where these words are spelled out as za-bar and si
par-ru. A next question is, how does the pronunciation Izabarl "derive" from the three
signs ud-ka-bar? In the word nidba, for instance, there was no obvious way to phonetically
relate the pronunciations Inindabal or Inidbal to any pronunciation of the individual signs;
that is, the word was more than the sum of its parts.
Since one of the three signs forming the word for "bronze" is the bar-sign, it seems
reasonable to assume that Izabarl derives phonetically from these three signs. B ut how?
One possibility might be to read ud-ka as zax . This type of approach is favored by many
Sumerologists, who try to make the writing system better fit Sumerian pronunciation.
However, there does not seem to be any other, independent, evidence which would justify
positing a reading zax for this sign, and the standard sign-lists do not recognize such a
value.
However, a further complication must be introduced. In the earliest Sumerian, the
word for "bronze" is not, in fact, written ud-ka-bar. Rather, it is regularly written as
KAxUD-bar, that is, with a ka-sign containing an inscribed ud-sign, followed by the bar
sign. For example, in a royal inscription of king Uruinimgina of Lagash, the word for
bronze appears as: L .
It is difficult to explain such a writing. Some scholars have posited a reading for
KAxUD; this is accepted by some sign-lists, although with reservation. However, Fal
ken stein has pointed out evidence that KAxUD can be read as sil 9 . This reading is
accepted by the standard sign-lists. Perhaps, then, the word should be transliterated as:
si 19-bar. Armas Salonen, in fact, transliterates this word as si l9 -bar for the older period,
but as zabar for the "nachsumerische" period ( 1961 : 108). In this interpretation, which is
probably correct, si l9 -bar represents an older pronunciation of the word. B ut the same
cuneiform signs continued to be used, when in spoken Sumerian the word had changed to
Izabar/.

- Writing system

The traditional interpretation of the nin-sign ( Bt3

) is that it represents the sign for
"woman" (munus, ), followed by the sign for "clothing" (tug, B ). That is, the sign is
a woman wearing a (special) kind of clothing, to mark her elevated status. This
interpretation has been questioned by Robert Biggs, who says that in the earliest Sumerian
texts, the tug-sign is different from the component forming the second half of the nin-sign;
1 26 Lesson 1 1

this is true to some degree even in Old Babylonian texts. This leaves the origin of the nin
sign up in the air.

- Overhanging vowels

The problem of the overhanging vowels has generated a good deal of polemics in the
course of Sumerological studies (the latest discussion is Yoshikawa 1980). This over
hanging vowel seems to be mostly la/, although individual cases of overhanging le/, liI,
and lul have been posited. Probably most Sumerologists accept their existence, but there
are dissenting opinions.
In Falkenstein's view, the readings in lal represent older forms of S umerian words.
That is, at one time these words were pronounced with a final la/. The word for "king",
for example, was originally *Ilugala/. At some (prehistoric) point, these final lals were
dropped; the word for "king" became Ilugal/, but the sign used to represent this word
could be used for either the newer value Ilugal/ or for the older value *Ilugala/. Since
both values co-occurred, a scribe could write "his king" as either lY.gal--ni, or lY.gala-ni,
both representing /lugalanil. (Presumably, the prehistoric form would have been *Ilugala
anil or *Ilugala-nil.)
Other Sumerologists question their existence; they see varying phonetic factors at
work. It was mentioned above, for example, that Yoshikawa believes that the genitive in
Ikl occasionally occurred after consonants, not just vowels; thus, "the son of the king"
might have been pronounced Idumulugalk/. However, Yoshikawa could not state any
general rules for the distribution of Ikl and lakl after consonants. For other overhanging
vowels (e.g., in the verbal system), he has other explanations.
In writings of the lY.gal-ni type, it has been posited that the lal was dropped:
*Ilugalanil ) Ilugalnil (this was Poebel's position). Without going into details, it can be
seen that such an explanation raises more questions than it answers, such as the co
occurrence of writings like lugill-ni and lugill -fl-ni in the same time and place.
Is there any independent evidence which justifies the view that certain signs contain an
overhanging vowel? Here the evidence varies, and is difficult to interpret. The fact that the
nominalized forms of the verb "to love" are written both ki-5!g5!2 and ki-g-g might seem to
indicate that this one sign can be read as lagl or laga/. However, it is also possible that
a
g is here a phonetic complement, and the form should be understood as: ki-5!g g ; or it
might be that these writings reflect phonological problems of particular roots ending in a
vowel.
The Akkadian lexical tradition is likewise ambiguous. Lexical lists do provide
readings with lal for some signs, but for the most part they do not (for example, they show
no evidence of a reading lY.gala). And, some of these readings with an overhanging vowel
may very well result from the Akkadian scribes encountering the same problems in the
writing system that we feel. These scribes may have anticipated some modem
Sumerologists, by generating readings in la/, in order to make the writing system more
closely fit the pronunciation.
Part of the problem may result from a misunderstanding of the nature of the Sumerian
writing system. Because the expression "malt house" is written -bappir, for example, and
 Lesson 1 1 127

not written -bappir-ra (for the assumed e.bappir.a(k , Falkenstein would say that the
second sign should be read as bappira. But it is easier to say that the Sumerian scribe felt
no need to write any indication of the genitive marker; such scribes were content to write
bappir, even if they pronounced it /ebappira(k)/. Falkenstein's school is an attempt to
make the writing system more closely resemble a transcription of speech, and this is not
how the writing system should be understood.
It is true, however, that there are other problems to be resolved. For example, "in the
land of Sumer" is normally written kalam-ma in the Gudea texts, for kalam.a. But once,
apparently, this locative phrase is written : kalam. Falkenstein would read this kalama.
Similarly, "on the tablet", written just dub, would be read by Falkenstein as duba. These
writings raise questions, but they may simply reflect an earlier period in Sumerian
orthography, when it was not necessary in general to write case-endings.

- Loan words

As just discussed, "bronze" is zabar in S umerian, siparru in Akkadian. The word for
"copper" is urudu in Sumerian, weru in Akkadian; they are usually spelled out in lexical
lists as !i-ru-du and -ru-.Y.. The ultimate origin of the word(s) is unknown. Both urudu
and wen1 may reflect one pre-Sumerian substrate word; they have even been connected
with the Indo-European word which ultimately appears in English as the adjective "red".
Curiously, in late Akkadian the word for "copper" also appears as urudG . Eduard Kutscher
says that "this 'Akkadian' word was artificially coined by Sennacherib's scribes (and used
only by them) from Sumerian urudu (= Akkadian wen1) 'copper'" ( 1982:225).

- History

Lamar is well-attested as an intermediary or intercessory goddess. She appears on Ur

III and Old Babylonian seals, introducing a worshipper to a higher god or goddess.
Because of this function, the name Lamar becomes almost a generic word for "protection".
Thus, there occur personal names of the type: Lugal-d Lamar-Rli I O , the king is my pro
tection". Borrowed into Akkadian, the word lamassu is glossed by the CAD as: "protec
tive spirit". Von Soden, however, has questioned the traditional derivation of lamassu from
Lamar.

- History

Other wigs have been found at various sites in Mesopotamia, although none of them
bear an inscription. For example, a wig of steatite was found at Uruk. Only 2.5 cm. long,
it was apparently designed to fit a statue (the dating is uncertain; before Nabonidus). More
recently, somewhat similar wigs have been found at Ebla.
128 Lesson 11

Text I I a
supplementary
 Lesson 1 2
This text was inscribed on the foot of a vase of marble. No photograph is available.

Sign-list and vocabulary

*" d rW dBiI-g-me Gilgamesh (DN, masc) ( = bi !)

J!' En-dim-gig Endimgig (GN ?)

(; ud U4) day
mu-sar-ra inscription

u ... ur to erase

rh1i f\ nam ...kurs <l9!s ) to curse

stv:=r bi
ib

-q b

f3=-:J da

Notes

dBU-ga-me This is now the preferred reading (or at least the more original reading) of
the name more familiarly known as "Gilgamesh". The apparent meaning of the name is:
"the old man (bi1-g is (now) a young man (mM)". It is not known when the change of
initial /b/ ) / g/ took place; an Old Babylonian omen text has the spelling dGe-el-g.
Some read the last sign of the name as mes.

mu-sar-ra sar is "to write". sar-ra sar.a, the passive participle: "something written".
=

mu has many meanings (e.g., "name"); mu.sar.a. is something like "a written text", or
"inscription". It was borrowed into Akkadian as musarU, translated by the CAD as: " I .
object bearing an inscription, 2. inscription".

u . .. ur This is a compound verb. u means "hand", and ur means something like: "to
move or drag (something)". u is thus the historical patient of ur: "to move the hand over",
i.e., "to erase". The verb takes its complement in the locative case. (The Akkadian

1 29
130 Lesson 12

equivalent is pasatg, glossed by AHw as: "tilgen, ausloschen".) ur also enters into the
formation of other compounds, for example: gis ... ur, "to harrow" (literally, "to drag wood
(over the ground)".
Some believe that the root ended in lul, and so it i s also transliterated as su ... uru 12'

nam ... kurS nam is apparently the (historic) direct object (patient) of kur5 ; the meaning is
something like "to cut a decision (against)". It usually takes its complement in the
comitative case. The word is further discussed below.
 Lesson 12 131

Text 1 2
132 Lesson 12

Transliteration Transcription Translation

1 : dBi I-g-mM Bilgame For Gilgamesh
2: En-di m-gigki Endimgig.(ak?) of En dimgig,
3: lugill--ni lugal.ani. (r) his king -
4: Ur_dNammu Umammu Ur-Nammu,
5 : nitab-kalag- nitabkalaga the mighty man,
6 : -.!! ' Ski-ma
I gI- Unm lugal. Urim.a(k) the king of Ur,
7: lugal-Ki-en-gi-Ki-uri-ke4 lugal. Kiengi. Kiuri.k.e the king of Sumer and Akkad -
8: ud _dNanna ud e.Nanna.(k).0 when he built the temple of
Nanna,
9: mu-du- mu.(n.)du.0.a.a
10: nam-til-Ia-ni-e nam. til.anUe for the sake of his long life
1 1 : -mu-na-ru a.mu.na.(n.)ru he dedicated a votive offering.
12: lu mu-sar-ra-ba lu musara. bi.a May Gilgamesh curse
the man who erases
1 3 : u-bi -i b-ur- u.bi .b. ur.e.0.a.(d) this inscription!
14: dBiI-g-me- Bilgame.e
1 5 : nam-b-ba-da-kur) nam.be. ba.da.kur.e.0
Commentary
2. The meaning of this line is unclear. The first editor of the text read it as en D I M.GIGki ,
"lord of D I M.GIG", an otherwise unattested place name. However, the editor added the
comment: "I cannot understand Gilgamesh's epithet in line 2".
It has been pointed out that there is an apparent GN, En-dim-gigki , which occurs at
least twice in Sumerian texts. There, the en-sign is apparently an element of the place
name, not the word "lord".
Presumably, lines 1 -2 form a genitive phrase: "Gilgamesh of Endimgig". This is,
curiously, the same construction found in one of the other occurrences of the GN: dNin
ubura EN.DI M.GIGki , translated by J. van Dijk as: "Ninubur von (?) EN.D I M.GIGki ".
A parallel to this construction (DN and GN in a genitive phrase) occurs in Text 17:
dNanna Kar-zid-da lugal-ki-.M-M.-ni-ir, "To Nanna ofKarzida, his beloved king".
In any case, one might have expected a final g-sign, to express the lal of the genitive
element.
8. ud introduces a subordinate, temporal clause: "when he built the temple of Nanna".
Sumerian does not have many different kinds of subordinate clause formation. The most
common type is a temporal clause. As is frequently the case in Sumerian, it is fairly easy ,to
recognize the surface form of the construction, but it is a little harder to understand the
grammar behind the written form.
The simplest temporal clause consists of: ( 1 ) a relative marker; (2) a relative clause
nominalized in .a, which stands in apposition to the relative marker; (3) a locative case-
 Lesson 12 133

marker in .a.
The most common relative marker is the noun ud (literally, "day"); others also occur.
Occ asionally the relative marker is deleted.
The essence of the relative clause is: "He built the temple of Nanna". This
independent sentence would be: e.Nanna.(k).0 mu.(n.).du.0. To form the relative clause,
this independent sentence is nominalized in .a. It is then placed in apposition to the relative
marker ud: ud [e.Nanna.(k).0 mu.(n.)du.0].a.
It is instructive to compare this relative clause with the relative clause in Text 7. In that
text, "the one who built the temple of Enlil" was expressed as:
lu [e.Enli1.a(k).0 i.n.du.0].a.
The construction here is much the same, but the relative marker is ud instead of lu:
ud [e.Nanna.(k).0 mu.(n.)du.0].a
Finally, the entire complex is put into the locative case, marked by .a (the second .a in
the transcription of line 9).
The function of the locative in .a here is "on": "on the day that". It is also possible to
find other case-markers, such as the ablative case-marker ta ("from, since"), or the
terminative case-marker Se ("until").
A literal translation of this entire clause would be: "on the day that he built"; this
captures both the force of the Sumerian locative, and the force of the relative clause. In
more idiomatic English, however, one may say "at the time when", or simply "when":
"when he built the temple of Nanna, he made a votive offering".
In this temporal clause, the verbal chain uses the conjugation-prefix mu, and the
personal-affix .n does not appear in the script. In the relative clause in Text 7, the
conjugation-prefix was 1. and the personal-affix .n appeared in the script. It is not easy to
understand the reasons for such alternations.
Although the transcription indicates both the .a of the nominalizer and the .a of the
locative case-marker, it is reasonable to assume that some kind of vocalic contraction took
place; one never finds two /als in the script in such a construction.
Lines 8-9 form the subordinate clause; lines 10- 1 1 form the main clause. In Sumerian,
the subordinate clause regularly precedes the main clause.
1 2- 1 5 . These lines express a curse; essentially the same wording occurs in other Su
merian inscriptions.
Lines 12- 1 3 are a relative clause, serving as the complement of the verb in line 15. Its
meaning is: "the one who erases this inscription". The relative marker is lu; the relative
clause is marked by the at the end of line 13.
Su ... ur is a compound verb. bi is another conjugation-prefix, in addition to the mu and
1 already seen. It is almost always written with the bi-sign. It will be further discussed
below.
All the verb forms seen up to this point have been in the gamn! aspect, used to express
action in the past. The verb form in this line, however, is being used to express a future
value: "whoever will erase", or "whoever might erase". Therefore, it is put in the man1
aspect.
It is common to speak of the "gamtQ-root" and the "marfi-root". The man1-root is
1 34 Lesson 12

fonned from the bamt!!-root in several different ways. The fonnation used for any par
ticular verb is lexical; that is, it is not predictable. It is not clear exactly how many classes
of maru fonnation exist. Yoshikawa has established three different classes:
(1) "Reduplication" class. The man1-root is fonned by (graphically) reduplicating the
bamt!!-root. Thus, "to return": bamt!!, g4; maru, g4-g4.
Roots of the pattern eve seem always to lose their final consonant when reduplicated.
Thus, "to place": bamt!!, gru:; man1, .M.-.M (always written with the g!!-sign).
(2) "Alternation" or "replacement" class. An entirely different root is used for the
maru. This root is non-predictable from the bamm-root. Thus, "to speak": bamm, .d!!g4 ;
maru, . dug4 and are two entirely different signs. However, there are cases where a
bamt!!-root and a maru-root will be written with the same sign. Thus, "to go": bamt!!, gin;
maru, duo The gin-sign and the du-sign are the same! In such cases, it is only the gramma
tical context which indicates whether the sign is to be read as the bamt!!-root or as the
maru-root.
(3) "Affixation" class. This is fonned by addition of .e to the bamt!!-root. This .e is
variously referred to as the "maru-element", the "maru-affix", or the "man1-suffix". Thus,
"to build": bamt!!, du; maru, du-. This class is the most common fonnation of the man1.
It is the fonnation used with the verb in Text 1 2, ur: ur.e. Here, however, the maru-suffix
has contracted into the nominalizer .a, and so it does not show up in the script.
In general, the particular maru fonnation for any specific verb is not always known.
And, some roots fall into two (or even all three) classes. As mentioned above, ur is a
member of the affixation class, but reduplicated man1 fonns also seem to occur. In later
Sumerian, combinations of these classes sometimes occur; e.g., a reduplicated root
followed by the maru-suffix. These cases have not all been explained.
The use of the maru, instead of the bamt!!, entails rather complex changes in Sumerian
morphology, particularly in the distribution of the personal-affixes. For a transitive verb:
in the bamm, the personal-affix slot before the verbal root cross-references the agent, and
the personal-affix slot after the verbal root cross-references the patient. In the maru,
however, it is j ust the opposite. The personal-affix slot before the verbal root cross
references the patient, and the personal-affix slot after the verbal root cross-references the
agent. For example, "The king built the house" is:

(1) lugal.e e. mu. .du.
But, "The king will build the house" is:

.s: *'
(2) lugal.e e.0 Lb.du.e.0
In (2), the ergative case marked in .e is cross-referenced by the .0 after the maru-
suffix. The absolute case marked in .0 is cross-referenced by the .b before the verbal root.
.b is used here to cross-reference inanimate antecedents; .n is used to cross-reference
animate antecedents.
That is, the case-markings on the nominal participants in the sentence are the same in
both aspects: lugal.e and e.0. However, the use of the personal-affixes is quite different.
Thus, in the verb fonn in this line, u.bi.b.ur.e.0.a.(d), the .e is the maru-suffix, and
the .0 cross-references the third-person agent. (This analysis of .e.0 is not universally
accepted, and a different view will be mentioned below.) The .b. before the root cross-
 Lesson 12 135

references the u, which is the (historic) patient of the verb ur.

Finally, the combination of bi . b. is quite frequent, and usually written with the bi -sign
followed by the i b-sign, as in Text 1 2.
The verb form may be summarized as:

u bi . b . ur . e . 0 . a
(1) (2) (3) (4) (5) (6) (7)
(1) nominal element o f compound verb
(2) conjugation-prefix
(3 ) personal-affix cross-referencing patient )
(4) verbal element of compound verb
(5) manl-suffix
(6) personal-affix cross-referencing agent (unexpressed)
(7) nominalizer, forming relative clause

mu-sar-ra-ba is for musara.bi.a, "on this inscription". . bi is a demonstrative, meaning

"this". It is suffixed to its noun. It is identical in form (and probably in origin) with the
third person inanimate possessive-suffix .
. a is the locative case-marker. Imusara.bil is in the locative case because it is the
complement of the verb u ... ur, which takes its complement in the locative. The Ii/ of Ibi/
has contracted into the lal of the locative.
To sum up, lines 1 2- 1 3 are a relative clause, meaning, "the one who erases this
inscription". This entire clause functions as the complement of the verb nam ...kur5 in line
1 5.
The compound verb nam ... kur5 most frequently takes its complement in the "comi
tative" case. The word comitative comes from the Latin word cum, meaning "with". It is
not easy to define or even to adequately describe all the uses of the comitative case, but it
often expresses ideas such as "along with": hl-da, "with the man". Besides this use, many
verbs in Sumerian take their complements in the comitative case. This usage is not usually
predictable, and must be listed in the dictionary.
The basic form of the comitative case-marker is .da, written: da.
There is a problem here, however, because there is no overt marker of the comitative in
the script: one would expect to find a da written at the end of line 1 3. One explanation for
its absence is to assume a phonological process similar to that of the dative case-marker .ra.
That is, Idal > Idl after a vowel, and word-final Idl is not written. However, in the royal
inscriptions of Gudea or Ur Ill, the comitative case-marker normally appears written as da,
even after a vowel; cases where an expected da does not show up in the writing only occur
in earlier Sumerian.
The problem of the presence or absence of the comitative case-marker in this line will
be further discussed below.
The scope of the assumed comitative case includes all of lines 1 2 and 1 3 :
[Iu muara.bi.a u.bi.b.ur.e.0.a] .(d).
1 4. The agent is marked in .e. The writing is morphemic; there is no attempt to graphically
reduplicate the final IU of dBi I-gil,-mM. Such morphemic writings are especially common
136 Lesson 12

with proper names; the name was felt as a unit, and the grammatical case-endings added
directly to the complete unit.
1 5 . b is a form of the "desiderative" or "precative" modal-prefix. As discussed in Lesson
1 , the very first element in the Sumerian verbal chain is an optional modal-prefix; there are
several of these.
The desiderative mood expresses wishes and indirect commands in the third person:
"Let him/them, may helthey", etc. The basic form of the desiderative is Ibel (written b);
before the conjugation-prefix ba, it regularly becomes Ibal (written b).
One of the more difficult questions in Sumerian morphology is the nature of the root
used after particular modal-prefixes. Some modal-prefixes use the bamm-root, others use
the man1-root, and still others use both, under conditions which are not always clear. A
second problem is the use of the personal-affixes. In some moods, the pre-verbal root slot
cross-references the agent. In others, it cross-references the patient. (This differentiation is
apparently irrespective of whether the bamt!!-root or the man1-root is used.)
With transitive verbs, M. is regularly construed with the mam-root; with intransitive
verbs, it is construed with the bamm-root. kurS is a member of the affixation class, and so
the form is: kurS -' The writing is morphemic in this line; in other cases the final consonant
of the verbal root may be graphically reduplicated (e.g., kurS -re).
In the desiderative, the agent is cross-referenced by .0 after the maru-suffix.
As with all compound verbs, the first element of the compound (here, nam) precedes
the entire verbal chain, including the modal-prefix.
ba is a conjugation-prefix not seen previously. It is usually assumed that it is related to
the conjugation-prefix bi , an example of which occurred in line 13. It is further discussed
below.
da is the dimensional-prefix which cross-references the comitative case. Here it cross
references the comitative case marked by the presumed .da at the end of line 13. The
original meaning of this compound verb may have been something like "to cut a decision
against".
To summarize the verb phrase, nam . . . kurS is a compound verb, with nam the (his
toric) patient. b is a form of the modal-prefix for the desiderative; with transitive roots, it
uses the maru form of the root. ba is a conjugation-prefix. da is a dimensional-prefix,
cross-referencing the comitative. The verb form may be diagrammed as:

nam be ba da kur . e 0
(1) (2) (3) (4) (5) (6) (7)
(1) nominal element of compond verb
(2) modal-prefix
(3) conjugation-prefix
(4) comitati ve dimensional-prefix
(5) verbal element of compound verb
(6) mam-suffix
(7) personal-affix cross-referencing agent (Gilgamesh).
 Lesson 12 137

Discussion: structure
The structure of this text is:

[Bilgame.Endimgig.(ak), luga1.ani].(r) benefactive

[Urnammu, nitab.kalaga, luga1.Urim.a(k), agent
luga1.KiengLKiuri.k].e
[ud e.Nanna.(k).0 mu.(n.)du.0.a].a time
[nam.ti1.ani].e purpose
a.mu.na.(n.)ru verb
[Iu musara.bLa u.bi .b.ur.e.0.a]. (d) accompaniment
[Bilgame] .e agent
nam.be. ba.da.kur.e. 0 verb

- Phonology

The second element of the compound verb nam... kur5 is variously transliterated as
kU5 ' kud, and kuru 5 (the latter with an overhanging vowel). The Idl - Irl alternation helps
show that the (amissable) Auslaut of this root was probably some kind of Irl sound which
did not exist in Akkadian. The Akkadian writing system sometimes reproduced it by Irl,
and sometimes by Id/. Modern Sumerologists sometimes transliterate it as dr, cIf, or t. It
seems to occur as the last consonant in some dozen or so verbal roots; another instance
occurs in Text 17. Its presence as the first consonant or medial consonant of a root is much
harder to detect.
The kur5 -sign can also be read tar. Confusingly enough, there appear to be two
different verbs: nam-kur5 meaning "to curse", and nam-tar, "to decide the fate ofIfor", "to
'
decree a destiny for". The boundary between the two expressions is, however, sometimes
unclear, and occasionally transliterations are less than precise in differentiating between the
two; both expressions are sometimes found transliterated as nam-tar.

- Moods

The morphology of the moods in Sumerian is quite complex. The single most
important work to unravel them is Dietz Otto Edzard 197 1ff; this is a series of articles
which it pays to keep close at hand. Some of Edzard' s conclusions were modified by
Burkhart Kienast ( 198 1 b).

- Conjugation-prefixes

In the paradigms presented in this book, model verbs in the bamt!! are generally cited
with the conjugation-prefix mu, and those in the maru are generally cited with the con
jugation-prefix 1. While this does represent the most common distribution (at least in the
Ur III royal inscriptions), it is also possible to find verbs in the bamt!! with the con
jugation-prefix 1 and verbs in the maru with the conjugation-prefix mu (although this latter
is rather rare).
It is sometimes stated that the conjugation-prefix ba "represents" bi with an additional
locative marker of some kind. It is difficult to say whether or not such a statement is
138 Lesson 12

correct. It is a fact, however, that when the conjugation-prefix ba is present, there is

frequently a locative phrase somewhere in the sentence. In purely synchronic terms, bi and
ba occupy the same slot. At least in the Ur ITI texts, they do not co-occur (that is, they do
not appear together in one verbal chain). It has, however, been claimed that there are cases
of ba and bi co-occurring in the somewhat earlier Gudea texts.
The conjugation-prefix bi differs from mu, i, and ba in that the only dimensional
prefix which can follow it is the locative dimensional-prefix .ni. The reasons why are
unsure.
Some scholars believe that writings such as bi-ni should be read as bi-i, with two
conjugation-prefixes. And as was said above, it has been speculated that ba and bi may co
occur in the same one verbal phrase in the Gudea texts. Such interpretations, which state
that it is possible for more than one conjugation-prefix to co-occur within one verbal
phrase, run counter to the view presented in this book, that only one conjugation-prefix can
so occur.

- Comitative - patient

In the analysis given here, lines 12- 1 3 are in the comitative case. However, the
apparent absence of the case-marker da does give one pause. Some people believe that
lines 12- 1 3 are the direct object of nam ...kurS . The comitative dimensional-prefix would
represent what Poebel called the "erstarrter Gebrauch des Infixes", that is, "frozen use of
the [dimensional] infix". In Gragg's study of the dimensional-prefixes in Sumerian literary
texts, he found a large number of instances where a comitative dimensional-prefix occurred,
without any corresponding comitative case-relationship.
This is not impossible. However, there is a more general issue here. In the compound
verb nam...kurS nam is apparently a (historic) direct object (patient) of kurS . Now, it is
'
usually believed that Sumerian does not permit two patients in one sentence. In cases
where one might expect two patients, one of them will be expressed through one of the
adverbial cases. In the immediately preceding lines, for instance, the English translation
was "to erase this inscription". The direct object (patient) in Sumerian was u; the direct
object in English, "this inscription", was expressed via a locative: "to move one's hand
over". If it is true that Sumerian does not tolerate two patients, and if it is also true that this
rule applies to historic direct objects (patients) of compound verbs, then lines 12-13 must
be marked by another case. Given the presence of the dimensional-prefix da in the verbal
chain in line 15, this would most likely be the comitative case.
The argument that Sumerian does not permit more than one patient in a sentence is
based on general linguistic theory, and on empirical observations in Sumerian. Some
linguists would say that no language has more than one patient (at least in the deep
structure); if two seemingly occur, one must be in an adverbial relationship. However, it is
not clear if such a constraint would apply to compound verbs. Even though the first
element of many compound verbs is historically the patient of the verb, it is not always sure
if it functioned as such in historic Sumerian.
 Lesson 12 139

-Conjugation
Following is the paradigm for the mam of the transitive verb in the singular. The
model used is sar, which is a member of the affixation class. The conjugation-prefix used
here is 1.
first person singular 1-sar-re-en Lsar.e.en
second i-sar-re-en Lsar.e.en
third i-sar-re Lsar.e.0
The first and second persons singular are identical in form. The final 1nl often does
not show up in the writing.
If the root ends in a vowel, there is frequent assimilation of the Iel of the marO-suffix
into the vowel of the root.
Some Sumerologists analyze the morphology of these endings as: .e.n, .e.n, .e.0.
That is, the first and second person markers are .n, not .en. This is a thorny issue, which
cannot be resolved here.
For verbs of the reduplicating class (the model is gar, "to place"), the forms are:
first person singular 1-R-g-en Lgaga.en
second l-g-g-en Lgaga.en
third i-g-g Lgaga.0
This interpretation of the maru of verbs of the reduplicating class is essentially that of
Yoshikawa. However, all such reduplicated forms end in a vowel, which is subject to
contraction with the lel of the ending .en. This means that in texts, such forms as the
following are encountered: i-M-M--en, i-gfi-gfi-an, i-gfi-M-, etc. Other Sumerologists
have argued that such writings indicate that Yoshikawa's analysis is incorrect.
The marker .0 in such marO forms as Lsar.e.0 has been treated here in two ways.
First, it was called a personal-affix, cross-referencing the transitive subject. Second, it was
called a marker for the third person. As discussed in Lesson 1, these are not contradictory
interpretations, but are rather two ways of saying the same thing.
- marO formation
It was Yoshikawa who established the three clases of mam formation discussed
above. Edzard has expanded this into five classes:
(1) "Unchanging": The hamtu-root and the maru-root are the same. This cor-
responds to Yoshikawa's "affixation" class; Edzard does not believe that .e is a marO
marker.
(2) "Reduplicating": Same as Yoshikawa.
(3) "Root-varying": The two roots are different, but similar phonetically: "To
approach" is te in the bamn!, but teg in the mam and teg are the same sign).
(4) "Replacement": Same as Yoshikawa.
(5) "Irregular": These do not seem to fit nicely into the other four categories.
140 Lesson 12

It is not yet clear whether Yoshikawa's or Edzard's classification scheme is to be

preferred. As progress in Sumerology is made, it is probable that more classes, and finer
subdivisions within these classes, will have to be made. Yoshikawa himself has indicated
that his scheme needs to be expanded; Kienast ( 198 1b) has suggested some modifications
to the scheme of Y oshikawa and Edzard. In this book, Yoshikawa's system of
classification into affixation, reduplication, and alternation classes has been followed,
because this classification scheme works well for the Ur III royal inscriptions.
- Roots
As may have been inferred above, it is only in the last ten or fifteen years that the
morphology of the maru has become somewhat clear, thanks primarily to the work of
Yoshikawa and Edzard. Even now, however, there remain thorny problems. Many
difficulties are occasioned by the fact that in some cases the same sign stands for two
different roots, one for the bamt!!-root, and one for the mara-root (such as girr/du men
tioned above). The situation is still more complicated, however; certain roots appear to
have separate forms for singular subjects and for plural subjects, in both the bam!! and the
mara; this produces at least four different roots. There appear to be cases where certain
roots have a bam!!-singular root, a mara-singular root, and a different plural root used for
both the bamt!! and the mara. In at least one case there appears to be a root which has
different forms, depending on whether the object is singular or plural.
This complexity is to be expected; other languages of the world show such diversity in
morphology. As progress is made, more such cases will be identified.
- Moods
There has been much discussion about the precise etymology of the Akkadian
grammatical terms bam!! and maru, and even more discussion about the distinctions which
are marked by these two terms. The difference in function between the bam!! and the mam
has been variously seen as a difference in tense, or a difference in aspect, or a difference in
Aktionsart. In the Ur III royal inscriptions, they seem more tense-like than aspect-like; the
bamt!! is regularly used for past action, and the man} for future action. However, when
dealing with more complicated texts, especially literary texts, such a single binary
distinction will not work. The fact that certain modal-prefixes require either the bam!! or
mam, seemingIy regardless of tense, complicates the issue.
-Ergativity
The term split ergative has been applied to Sumerian, because the personal-affixes
behave in an ergative way in the bam!!, but not in the maru; in the man}, they behave in a
nominative-accusative way. Consider the following sentences:
( 1) The king built the house .
lugal.e e. fi2 mu.n.du.@
.:t .....

(2) The king went.

lugal. fi2 mu.gin.
 Lesson 12 141

(3) The king will build the house.

'" ;)-
lugal.e e. .du.e.0
(4) The king will go.
lugal. Ldu. (Q
In (3), the direct object is cross-referenced by the personal-affix in the pre-verbal root
slot. In (4), the subject of the intransitive verb is cross-referenced in the post-verbal root
slot. Since the direct object in (3) and the intransitive subject in (4) are not cross-referenced
in the same manner, they cannot be considered to function in an ergative way, but rather in
a nominative-accusative way.
The case-markers in (3) and (4) - and in ( 1 ) and (2) - are the same; the difference is in
the way that the case-markers are cross-referenced in the verbal phrase. In ergative
languages which lack a case-system, ergativity only shows up in the cross-referencing
system. Discussing a Mayan language called Sacapultec, for instance, Du Bois says: "As
in all Mayan languages, the ergative patterning of Sacapultec morphology is entirely in the
verbal cross-referencing inflection; nouns are not case-marked for grammatical relations"
(1985:809).
Split ergative languages help demonstrate that the ergative - accusative distinction is
more of a continuum than as a simple dichotomy. Du Bois has suggested that split
ergativity is motivated by pragmatic or discourse pressures, forcing a re-alignment of the
principal constituents of a sentence.
The explicit statement that Sumerian is a split ergative language has been particularly
advanced by Michalowski. Not everyone is in agreement with these views; some scholars
believe that there are too many apparent exceptions which cannot yet be explained.
-Reduplication
Reduplication plays many roles in Sumerian. In this Lesson, reduplication is one of
the devices used to derive maru-roots from bamm-roots. A second common use is often
referred to as "free" reduplication, or "bamm" reduplication. This consists of reduplication
of the bamm-root; for example, l- gm:-gar. This is not a case of mara reduplication, because
the mara of gar is i-M- g.
The functions of free reduplication are not all clear. In many ways, it appears to be the
functional equivalent of the D-stem ("Intensive") in Semitic. Just as it is not easy to
categorize all the uses of the D-stem in Semitic, so it is not possible to easily categorize all
the uses of free reduplication. Edzard's preliminary classification includes such things as:
stressing of plural or totality of subject or object; plurality of occurrences; distributive
relations; etc. In bilingual literary texts, reduplication is often translated by the Akkadian
Gtn ("Iterative") stern.
Free reduplication is not uncommon in Sumerian; for example, it occurs frequently in
Gudea and in Old Babylonian literary texts. There appear to be no cases among the Ur III
royal inscriptions, perhaps simply because of content; plural objects do not appear to be
mentioned.
Certain verbs seem to have become lexicalized in the form of a reduplicated root. For
1 42 Lesson 12

example, .lli4 -ur4 ' "to look for and gather up", almost always appears this way.
Presumably, this is because it is almost always used with a plurality of objects. Such a
lexicalization has a parallel in the Semitic languages, where certain roots are lexicalized in
certain stems. For example, the Akkadian buu, "to look for", only occurs in the D-stem.
In context, it is often very difficult with certain verbs to decide whether marii re
duplication or free reduplication is present. And if it is a case of free reduplication, it is
often not easy to see its function.
It is probable that reduplicated forms (of whatever kind) were phonetically reduced in
speech. There is evidence for this from texts written in syllabic orthography (Appendix 2)
and from unusual occurrences of syllabic writings within normal orthography. For
example, an expected biz-biz- appears once in Gudea as: bi-bi-ze. Few details of such
phonetic reduction are understood.
- Origin of cases
It has been speculated that the Sumerian case-endings (some of them, at any rate) were
originally nouns. In particular, it has been claimed that the comitative case-marker da is in
origin the same da meaning "side", seen in the expression "king of the four quarters",
lugal-an-ub-da-limmuba. This is not impossible, but it is harder to find an etymology for
the other case-markings.
- Research in Sumerian
It is sometimes fairly easy to understand the meaning of a Sumerian text, and even
relatively easy to describe, on surface terms, what we see, but it is much more difficult to
understand exactly what is happening. For example, based on context and on parallels in
Akkadian, Phoenician, and Aramaic texts, all Sumerologists would understand the last lines
of this text to mean, "May Gilgamesh curse the man who erases this inscription",
regardless of the presence or absence of a comitative case-marker, regardless of the
presence or absence of a dimensional-prefix, regardless of the distribution of the personal
affixes, etc. However, unless such details are well understood, it is much more difficult to
figure out the meaning of really complicated passages.
This text also illustrates the problems encountered in doing research in Sumerian. An
obvious question which arose when discussing the presence or absence of a comitative .da
at the end of line 13 was, "How is this verb construed in its other occurrences? Does it use
the comitative or some other construction?" Unfortunately, without an up-to-date Sumerian
dictionary, such questions are not easy to answer. One can look at the existing dictionaries,
or texts with glossaries, but without a painstaking examination of many sources, it is
impossible to be sure that all instances of any particular word have been found.
-Curse-formulae
It is not uncommon for votive inscriptions to be provided with a curse-formula.
Typically, the first part of an inscription will form a straight-forward text; the curse is
tacked-on at the end. In the Ur III texts, only a limited number of curse-formulas occur; the
next occurrence is in Text 15.
 Lesson 12 143

- History
The historical Gilgamesh was the fifth king of the First Dynasty of Uruk, which falls
within the Early Dynastic IT period (about 2700-2500 BC). No inscriptions of his are
preserved, or contemporary references to him, but there are a few inscriptions of his
approximate contemporaries.
The first attestation of him is in a god-list from Fara, where his name is written: dBiI
PAP-g-mM. It is difficult to say exactly how these signs represent the name /BilgameS/;
his name is spelled several different ways in the course of Mesopotamian tradition.
Gilgamesh seems to have been very popular with the Ur III kings; according to Jeffrey
Tigay, "The kings of Ur ITI regarded Gilgamesh as something like their personal god"
(1982: 1 3 n.50). Michalowski speculates that the "Gilgamesh stories were made part of the
school curriculum during the Ur III period" (1987:66). In his hymns, Shulgi refers to
Ninsun, the mother of Gilgamesh, as his own mother, and he refers to Gilgamesh as his
brother.
144 Lesson 12

Text 1 2a
supplemen tary

Another door socket.

r r
1T Tt:>
i * ZZ
m <T : :}
I

5
z t: r:J
l' f
r m
2ffi -<r EU IUIl
m:A(/J t=4;> *
f tr>
*-7 t:>
2
m tlti
W f < ma tt> t
10

H H
 Lesson 12 145

Note.

-dur-an-ki The name of the ziggurat at Nippur. Etymologically, "the temple, the bond
of (= between) heaven and eanh", E dur.[an.kiJ.k.
Text 12a was found outside of a controlled archaeological COnlext. The following
photograph is of a door socket bearing another copy of the same inscription, found in situ:

THE INSCRIPTION THAT IDENTIFIED THE TEMPLE Of' INANNA

A. door socket. this granite block bore a door post in its cavity. With the discovery o f the thick buttressed
wall we had only the outside of a monumental building. To get inside. two plckmen dug a five.yard square
shaft beside the inner wall face. Walls and floors were soon located, but the room was larger than the shaft.
TIle few objects found did not identify the building. To learn more about the area with a minimum of
effort, we tunneted atong one wall for twenty feet, then another (or thirty feet to a corner where there was
a doorway. We dug through this, first finding a disappointingly uninscribed door socket, then. in a brick
box tower down, another with cuneiform wedges. The writing is that of king Shulgi, second ruler of the
Third Dynasty of Ur. commemorating his rebuilding of the temple of Inanna. With this, we had the cult
spot of Inanna at Nippur.
 Lesson 1 3
This text is a standard inscription of Amar-Sin, the son and successor of Shulgi; he
ruled from 2046 to 2038 BC. The text exists in many copies; Text 13b is a stamped brick.
Sign-list and vocabulary

Zuen Zuen (DN, masc)

o* ;! Amar-dZuen Amar-Sin (PN)

Nibru Nippur (ON)

amar young bull

sag-us supporter, sustainer, patron

<m pad () to find, call, reveal

mu ...pad (w to propose

Notes

Zuen This seems to be another name of Nanna, although it is not clear why he had two
names. Jacobsen thinks that the term Nanna refers specifically to the god's role as the "full
moon", and Zuen refers to his role as the "crescent moon". It has also been suggested that
Zuen is the Akkadian equivalent of Sumerian Nanna; that is, they are two different names
for the same deity. However, there is no obvious Semitic etymology for Zuen.
The Akkadian equivalent of this DN is usually transcribed as either "Sin" or "Suen".
The Sumerian word was also borrowed into Akkadian as a common noun, appearing as
sInu, suenu, sinnu, and innu. It is glossed by the CAD as: " 1 ) the moon 2) crescent
shaped or semi-circular object".
The writing is discussed below.
Amar-dZuen Etymologically, "young bull of Zuen", amar.Zuen.(ak). The name of this
ruler is often transcribed "Amar-Sin" or "Amar-Suen", which are really Akkadianized
transcriptions.
Early scholars believed that this PN was Akkadian. The Akkadian equivalent of amar
is bOru. Therefore, the name appears in some older secondary literature as: "BOr-Sin", or
something similar. Almost all modern scholars believe that the name is Sumerian, although
there is really not much evidence to prove this.

147
148 Lesson 13

Nibru One of the more ancient and important cities in Mesopotamia; Kramer has called it
the "spiritual and intellectual center of Mesopotamia". Jacobsen has said:
From the very beginning of historical times Nippur and Enlil were recog
nized as an undisputed source of rule over Sumer as a whole, and kings of
Sumer would derive their authority from recognition in Nippur rather than
from their own city and its city-god ( 1957:139).
The modern name of the site is Nuffar. It was the first tell to be excavated by
American archaeologists (the University of Pennsylvania, beginning in 1 887). Over thirty
thousand tablets were found, mostly in Sumerian, ranging from the third to the first
millennium BC. Large numbers of these texts are still unpublished. The vast majority of
our Sumerian literary texts are Old Babylonian copies found at Nippur.
The etymology and writing of the name are discussed below.
sag-us The exact etymology and meaning are not sure. , as seen previously, means
"head". us has several meanings. It often means "to lie against", "to lean against". (It is
used as a logogram to represent Akkadian emedu, "to lean against, to reach".) It may also
have a transitive sense, "to lift", i.e., "to support". Here it may be an active participle in .0,
with sag being its historic patient or incorporated object: "the one who lifts the head". The
usual translation is "supporter, sustainer, champion (of)". It is translated by Sollberger as:
"protector, patron (literally, '(he who) supports the head')".
 Lesson 13 149

Text 1 3 a

* D *r-@ :::[!]
.
Ir " -f
.._ -
. -.
.- .. .
<9 'J
r:li1 ,
*" r--'fi 111-
111- )ECf
<J-- <> @hj
11
-"4

Cif.- r -h1
firFl * J7J . . ..- r - -I
.
...
-
-
- -

i . . . .-

.I,' I , Ij (!!:>
I.b, /iil :> t:t:::J
- I --- ,I f'

' *
=

There is a certain amount of variation in the cuneiform signs in this text. The da at the
end of line 4 has an initial vertical, not seen in the da in line 9. Similarly, the sign read nitab
in line 7 has a vertical not seen in the sign read us in line 5.
150 Lesson 13

Text 1 3b
 Lesson 13 151

Transliteration Transcription Translation

1 : dAmar-dZuen Amarzuen Amar-Sin,
2: Nibruki_ Nibru.a roosed
Ey nlil
3: dEn-lil-Ie Enlil.e in Nippur,
4: mu-pad-da mu.pad.a
5: sag-us sagus. patron of
6: _dEn-lil-ka e.Enlil.(a)k.a(k) the temple ofEnlil,
7: nitab-kalag-g nitabkalaga the mighty man,
8: ..!! ' Ski -ma
I !@!I-Unm lugal.Urim.a(k) the king of Ur,
9: lugru-an-ub-da-lirnrnuz-ba lugal.anub.da.(k) limmu.bLa(k) the king of the four quarters.
Commentary
1. As was the case with Shulgi, the name Amar-Sin is preceded by the determinative for
divine names. There are thus two divine detenninatives in the line: the second is for the
DN Zuen, and the first is for the PN Amar-dZuen.
2-4. These lines form a relative clause modifying Amar-Sin of line 1 . mu means "name",
and pad is something like "to reveal". mu ...pad is a compound verb, meaning approx
imately "to propose". The underlying idea is that the name of Amar-Sin was proposed by
Enlil in the council of the gods, meeting in Nippur, to be the king of Sumer and Akkad.
Sumerian has two ways of forming relative clauses. The first has occurred several
times: it consists of nominalizing a complete independent sentence by .a, and then placing
this nominalized sentence in apposition to a relative marker. This is sometimes called the
"full" relative clause. For example, "He built the temple of Nanna" is: e.Nanna.(k).0
mu.n.du.0. "The one who built the temple of Nanna" is: lu.[e.Nanna.(k).0 mu.n.du.0].a.
In this example, the relative marker is logically the subject of the verb in the relative
clause: "the man who built", "the man who shall erase". However, oblique relations are
also possible, such as: "the temple which the king built", "the god for whom the king
dedicated a votive offering".
In a clause such as "the temple which the king built", "temple" is logically the direct
object of the verb. In such relative clauses in Sumerian, there is no overt marker indicating
this relationship. This clause could be expressed as: lugal-Ie mu-du-. The simplest way
to understand this construction is to think of it as: "[the temple] [the king built (it)]". In
English, there is no overt marker, except in the distinction between "who" and "whom":
"the man who built", but "the man whom he saw". In the classical Semitic languages, a
resumptive pronoun is used. For example, the Akkadian equivalent of "the temple which
the king built" is: bItu a arru ibmlu. Literally, this is: "the temple which the king built
it". Sumerian uses no relative marker of any kind (unlike the use of English "that, which",
or Akkadian a), and there is no overt marker for the direct object (unlike the Akkadian u).
The second way relative clauses are fonned in Sumerian is less understood in all its
details. It is sometimes called a "reduced" relative clause, and sometimes a "participial
152 Lesson 13

construction". It is fonned by deletion of the entire verbal prefix chain, and nominalization
of the remaining verbal root in .a (except in certain cases, not discussed here).
To express "Amar-Sin, whom Enlil proposed in Nippur" using the full form of the
relative clause, would be approximately: Amarzuen [Enlil.e Nibru.a mu.mu.n.pad.0].a.
(The first Imul is the nominal component of the compound verb mu ...pad: the second
Imul is the conjugation-prefix.) To express the same idea using the reduced relative
clause, the prefix chain is deleted: Amarsin [Enli1.e Nibru.a mu.pad.0].a, which is the fonn
in Text 13. The mu which remains is the nominal element of the compound verb, not the
conjugation-prefix. Since the nominal component of a compound verb is not part of the
prefix chain, it is not deleted.
Several English translations of this construction are possible: "Amar-Sin, whom Enlil
proposed in Nippur", "Amar-Sin, proposed by Enlil in Nippur", etc.
Both full and reduced relative clauses are common in Sumerian, but it is not known if
there are rules governing their distribution. Certain fonnulaic expressions tend to prefer
one construction, while other fonnulaic constructions prefer the other. For example, "the
man who built ... " always appears in these texts as lu in-du-, but "the man proposed by
Enlil" always appears in these texts as lu dEn-lil-Ie mu-pad-da.
5. sag-us _dEn-lil-ka sagus.e.Enlil.(a)k.a(k), "patron of the temple of Enlil". The geni
=

tive phrases seen till now have consisted of two nouns or nominal phrases. However, it is
also possible to have a genitive phrase consisting of three or four elements, such as "the
king of the temple of Nanna". Such genitive phrases are fonned by the addition of an extra
".ak" for each new element in the genitive phrase. For example, "the temple of Nanna" is:
e.Nanna.k; "the king of the temple of Nanna" is: luga1.e.Nanna.k.ak; "patron of the temple
of Enlil" is: sagus.e.Enli1.ak.ak. (Sequences of two ".ak"s are sometimes referred to as
"double genitives".)
Sequences of four nouns or nominal phrases (therefore, with three ".aks"), although
pennissable, are uncommon. Sumerian does not seem to tolerate a sequence of more than
three ".ak"s; if such a situation would arise, no more than three are used. More commonly,
a periphrasis of some kind is used instead.
The genitive phrase in this line is written sag-us-_ dEn-lil-ka. The expected la! of the
first genitive marker does not appear in the writing. One might have expected a writing
such as sag-us-_dEn-lil-Ia-ka, or some such. A similar writing occurred in Text 1 1 : bi-li
nam-munus-ka-ni, for bili.nam.munus.(a)k.ani. As discussed at length in that Lesson, it is
not known whether the problem is at the orthographic, phonological, or morphological
level. Falkenstein's school, for example, would read the lil-sign here as lila2 : sag-us-
dEn-lil!!z"ka.

Discussion: structure
Text 13 is a standard inscription, similar to Text 7. It consists entirely of a series of
appositives, serving as epithets to the name Amar-Sin in line 1 ; there is no finite verb form:

Amar-Sin,
proposed by Enlil in Nippur,
 Lesson 13 153

patron of the temple of Enlil,

the mighty man,
the king of Ur,
the king of the four quarters.
-Orthography
The name Zuen is composed of two signs, the en-sign and the zu-sign. The two signs
are always written in this order. However, there is a fair amount of evidence which shows
that the zu-sign was actually pronounced before the en-sign; that is, this name was
pronounced something like /zuen/. For example, the Akkadian word sInu (discussed
above) is a loan-word from Zuen. In the bilingual texts from Ebla, the Sumerian version of
this divine name appears once as En-zu, but twice as En-zi. In all three cases, the Eblaite
equivalent is Zu-i-nu (read by some as Su-i-nu). Also, there are Akkadian personal names
formed with the Akkadian version of the name, which appear in Hebrew, Greek, Latin and
later English transcriptions; thus, the name Sin-ahhe-erIba ("Sin has replaced my [dead]
brothers") appears ultimately in English as "Sennacherib". There are, however, some
unusual spellings in later Sumerian texts which may indicate that at times the name Zuen
was "read as written", that is, read as /enzu/ and not as /zuen/. These late writings may
result from misunderstandings of scribes.
It is not known why this order of signs is used; a similar phenomenon is discussed in
Lesson 14. In very early Sumerian, it was possible for signs within a line or case to be
written in a rather free order, not always corresponding to the order of signs as they were
read or pronounced. It is possible that writings such as En-zu represent survivals from this
period.
It is difficult to decide how to transliterate such spellings. Some Sumerologists
transliterate sign-by-sign; thus, En-zu in this case. This practice is not common in modern
day transliterations, and is regarded as somewhat old-fashioned. Other Sumerologists
follow a convention whereby the transliteration gives the signs in the order-as-read,
separated by a colon: Zu:en. Some Sumerologists will then follow this with the trans
literation of the signs in their order-as-written, usually in caps, within parentheses:
Zu:en(EN-ZU). The colon is sometimes used in this way when transliterating very early
Sumerian, when the order of signs in the line or case does not correspond to the presumed
order of pronunciation.
- Writing system
The Sumerian pronunciation of the name of the city of Nippur is known from lexical
lists, where En-mki is spelled out as Ni-ib-ru. Similarly, the Akkadian pronunciation of the
city name is also known from lexical lists, where it is spelled out as Ni-,ill-p!!-ru.
The Sumerian writing of the place name represents a not uncommon instance where
the writing system tells us nothing about the pronunciation of the place-name. The
etymology of Nibru is unknown; it is presumably a pre-Sumerian substrate word.
However, the city eventually became especially associated with the god Enlil. Therefore,
the name of the city was written with the same two signs as in the god's name, but followed
1 54 Lesson 13

by the detenninative for place: En-lilki . That is, the writing does not attempt at all to repro
duce the phonetic sequence INibru/. Rather, the Sumerian reader would understand the
written signs as standing for "the place associated with the god Enlil", that is, Nibru.
Transliterations of such place names (and of similar common nouns) vary. Older
practice tends to reproduce the basic value of the signs fonning the word: Ep._lilki . More
current practice is to use the name of the GN (assuming it is known) : NibrukI.
-Relative clauses
The tenn "participial construction" has been used to describe constructions such as in
lines 2-4, because a reduced verbal fonn with a nominalizer is fonnally identical with what
has been called here a passive participle; a reduced mu-n.-du, becoming du-, is fonnally
identical with the passive participle du-!.. In origin, in fact, passive participles are all
probably reduced relative clauses, in special syntactic environments.
The construction called here "reduced relative clause" is often referred to as the
"Mesanepada (or Mesannepadda) construction". (This name for the construction
apparently goes back to Falkenstein.) Me-an-ne-pad-da was the founder of the First Dy
nasty of Ur ("Ur I") sometime around 2550 BC. His name means "the young man whom
An chose", or "the young man chosen by An": Me.an.e pad.a. This is the minimal fonn
of the construction: a head noun (mM); an agentive marked in .e (An-; a verbal root
(pad); a nominalizer. (A few other names of the type "X-an-ne-pad-da are also known.)
Lines 1-4 of Text 13 are only slightly more complicated; Text 13 includes a locative phrase,
and also uses a compound verb (mu ...pad) instead of just pad.
Other examples of this construction occur in Gudea. A temple is referred to as: E
ninnu An-ne ki-gar-ra, "the Eninnu temple, which An established (literally, 'placed on the
ground', ki.a)", or "the Eninnu temple established by An".
The only study that deals specifically with relative clauses in Sumerian is by Gragg
(1972a). It was written for a non-Sumerological audience, and is by design short and
schematic, but has several useful observations. He points out that the syntax of relative
clauses, particularly reduced relative clauses, is not completely understood. Henri Limet
(1975a) has studied the parallel use of .a in participial and relative sentences.
- Standard inscriptions
Copies of this text have been found at several different sites, including Adab, Bad
Tibira, Eridu, Girsu, Isin, Kisurra, Sippar, Tell el-Lahm, Ur, and Uruk. All copies were
inscribed on bricks. This situation is not uncommon with standard inscriptions, which can
be found anywhere the ruler held sway, or where building-activity was conducted under his
aegIS.
- History
Shulgi apparently died of old age; he had ruled for some 47 years. His son, Amar
Sin, ruled only nine years, and not much is known of his activities.
 Lesson 13 155

Text 1 3c
supplementary

Another brick.
 Lesson 1 4
This text is a brick building inscription of Amar-Sin; it is essentially an expansion of
Text 1 3.
Sign-list and vocabulary

En-ki Enki (DN, masc)

abzu "apsu", water basin

Notes

En-ki The god of the subterranean waters, and also the god of wisdom. He was a son of
Nammu. His name apparently means "lord of the earth", en.ki.(k). There are spellings
which show that this name is a genitive phrase, not a noun-noun compound.
Why a god whose name means "lord of the earth" became associated with water is not
entirely clear; Jacobsen has spoken of "the role of water in fructifying the earth". It has
also been speculated that the element Ikil appearing in this name is a different word than
the word lkil meaning "earth", perhaps the same ki appearing in the compound verb
kig2
The Sumerian god Enki was equated with the Akkadian god Ea <E.-. The name of
the latter is of uncertain etymology; it does not inflect for case. In the bilingual lists from
Ebla, the Ebaite equivalent of Enki is written E-!!9 ' This would appear to be an inflected
form of the name, with the nominative case-marker. It has also been speculated that the
Akkadian writing E- and the Eblaite writing E-!!9 are actually phonetic spellings repre
senting a Semitic form something like Ibayyu/, "the living one". This idea is explicitly
developed by Cyrus Gordon (1987:19-20).
abzu This is composed of two signs, the zu-sign followed by the ab-sign. However, it is
known that the ab-sign was read before the zu-sign. That is, the word was pronounced
something like labzu/. The phenomenon is similar to that of the DN Zuen, which was
pronounced lzuen/, although written en-zu. In older transliterations it may appear as zu
ab.
The original use of this term was mythological. It referred to the subterranean fresh
waters, which the Sumerians believed lay below the surface of the earth. These waters fed
the wells, streams, rivers, marshes, etc. These waters were the special purview of Enki.
The term abzu was also used as the name of a large temple in Eridu, built to honor
Enki. Most of the work on this temple was done by Amar-Sin, although it was his father
who actually began the construction. This temple apparently stood over a fresh-water
lagoon.
The term was later applied to a cultic object, presumably some kind of water-basin
157
158 Lesson 14

used in the temple. At a number of sites in Mesopotamia, objects have been found which
archaeologists have identified with the term abzu. Several such objects have recently been
found at Ebla.
This word was borrowed into Akkadian as apsfi, glossed by the CAD as: " 1 ) deep
water, sea, cosmic subterranean water, 2) (a personified mythological figure), 3) water
basin in the temple".
It is possible that the Eblaite equivalent of Sumerian abzu and Akkadian apsfi appears
as -ba-si (Fales 1984: 1 84), but the interpretation is somewhat uncertain.
The English word "abyss" is thought to derive from this Sumerian word, via
Akkadian and Greek. It has been speculated that the word is not native Sumerian, but
rather derives from a substrate language.
 Lesson 14 159

Text 1 4

-t Jtt Tt
)-> z
S:j
ffiPT

'"* :f-:i
:d gg l
I

..*,- --M lJJ

t\ g !=: I
.M>
----

r!j\
160 Lesson 14

Transliteration Transcription Translation

1 : dAmar-dZuen Amarzuen Amarsin,
2: dEn-lil-Ie Enlil.e groosed
y nlil
3: Nibruki_ Nibru.a in Nippur,
4: mu-pad-da mu.pad.a
5: sag-us sagus patron of
6: _ dEn-li1-ka e.Enlil.(a)k.a(k) the temple ofEnlil,
7: lugal-kalag-gi! lugal.kalaga the mighty king,
8: ...! ' ki-ma
I !gilIl-Unm lugal.Urim.a(k) the king of Ur,
S
9: lugill-an-ub-da-limmuI ba-ke4 lugal.anub.da. (k) the king of the four quarters -
limmu.bi.ak.e
1 0: dEn-ki Enki for Enki,
1 1 : lugal-ki-g-M-ni-i!: lugal.ki.aga.a.ani.r his beloved king -
12: abzu ki-g-M-ni abzuki.aga.a.ani.0 his beloved apsii -
13: mu-na-du mu.na.(n.)du.0 he built.
Commentary
9. lugal-an-ub-da-limmuTba-ke4' Cf. line 9 of Text 13: l!!gal-an-ub-da-limmu2 -ba. The
difference between the two is the presence of the ergative case-marker in Text 14. It was
not present in Text 13, because there was no finite verb form in that text; rather, Text 13
consisted of a string of appositives. B u t i n Text 14, all the appositives are part of the
nominal phrase expressing the agent of the transitive verb in line 13. As stated in Lesson 1 ,
the nominal phrase to which the case-markers are attached in Sumerian can vary
considerably in size - all the way from a single noun, to long complexes such as this one: a
nine-line nominal phrase.
1 1 . The dative case-marker Irl is here expressed. Its occurrence in the royal inscriptions
of Ur-Nammu is unsure, and it does not appear in the inscriptions of Shulgi. At some
point in the reign of Amar-Sin, there was apparently a change in orthography, although the
motivation for this full writing is unknown. During the time of Amar-Sin there are wri
tings with the dative case-marker Irl expressed in the writing, such as here, but there are
also texts where it is not expressed.
The problem cannot be described simply in chronological terms. Even in the Gudea
texts, there are isolated instances of the Irl appearing in the script. For example, "to the
king" is normally written either l!!gal-il,.-ni or l!!gal-ni in Gudea, but l!!gal-ni-ir occurs at
least once. "For Gudea" is written: Gu-de-; the spelling Gu-de-i!-ar is also attested. Fal
kenstein, in his study of the Gudea inscriptions, could find no rules governing either the
morphology or the orthography of the dative.
It has been argued that the problem here is phonological, not orthographic. Poebel and
Falkenstein have suggested that in a writing such as l!!gal--ni for the dative, the original
 Lesson 14 161

Irl of the dative case-marker was completely lost, producing /lugalnil for the dative.
However, adoption of such a view entails rather baroque convolutions in explaining such
forms as h!gal--ni-ir. Falkenstein, in fact, speaks of "eine sekund11re Restitution" of the
dative case-marker Irl, reflecting a period when Sumerian was beginning to fall out of use
as a spoken language. This would mean that the original form was Irl, but then the Irl
was lost, and later it was "restored". However, such an explanation encounters strong
linguistic objections, and also historical objections. Also, it does not account for the rise of
explicitness seen in other areas of the grammar. In Text 16, for example, there occurs mu
na-an-du, the first occurrence in the body of texts studied here where the personal-affix
cross-referencing the bamt!!-agent is actually written.
1 2. Because of the ambiguity of the term abzu, it is not clear what the patient in this line
refers to: the temple built by Amar-Sin and his father, or to a cultic object within this
temple.

Discussion: structure
The structure of this text is:

[Amarzuen, Enlil.e Nibru.a mu.pad.a, agent

sagus.e.Enlil.(a)k.a(k),
lugal.kalaga, lugal. Urim.a(k),
lugal.anub.da.(k) limmu.bLak].e
[Enki, lugal.kLaga.a.ani].r benefactive
[abzu ki.aga.a.ani].0 patient
mu.na.(n.)du.0 verb

The first nine lines are essentially the same standard inscription seen in Text 1 3. The
rest is a straight-forward building inscription. The result is that this inscription begins with
the name of the king (the agent) instead of with the name of the deity (the benefactive). It is
as if the scribe began with a stock standard inscription, then tacked on a building
inscription.
There are two differences between Text 1 3 and Text 14. The first is in the order of
lines 2 and 3. In Text 1 3, the locative phrase within the reduced relative clause precedes the
agentive phrase, but in Text 14, the agentive phrase precedes the locative phrase. Text 14
uses the more usual syntax. In Text 1 3, there is presumably some emphasis on the word
"Nippur". The second difference is in line 7. Text 1 3 uses a title which occurred in several
other texts: nitab kalag-g. Text 14 uses lugill kalag-g,a.

- Orthography

On the one hand, writings which explicitly represent the dative case-marker Irl may be
viewed as part of the general process of the Sumerian writing system becoming more and
more explicit in its representation of phonological and morphological features. On the other
hand, it is hard to understand exactly how such a practice appeared - what motivated a
scribe, practicing by its nature a conservative craft, to write the Ir!?
It is not known when Sumerian began to die out as a spoken language; many scholars
162 Lesson 14

believe that such a process was already on-going during the Ur III period. If so, the
increase in explicitness in the texts - such as the writing of the dative case-marker Irl -
may be correlated with an increased need of the scribes for help in reading and writing
Sumerian. That is, as the scribes' knowledge grew more and more "shaky", there was a
need to write the morphemes down in an unambiguous way. At the same time, there may
have been felt a scholastic tendency to write all morphemes down. Similarly, in our own
scholastic tradition of transcribing S umerian, we are prone to write down full underlying
forms of morphemes.

-Ergativity

There are two ergative case-markers in this sentence: the .e in line 9, marking the
agent of the main verb in line 1 3, and the .e in line 2, marking the agent of the verb in line 4
(embedded in a relative clause). Potentially, this could cause a certain amount of confusion.
If one thought that the .e in line 2 marked the agent of the verb coming up in line 1 3, the
text would start to become rather confused. In practice, however, the formulaic nature of
these texts helps to prevent such confusion. In the spoken language, there were probably
features such as stress and intonation which helped obviate such problems.

- Textual problems

At the end of line 9, the autograph reads ke4 ; the ke4 -sign combines the Ikl of the
genitive marker with the lel of the ergative case-marker. This particular use of the ke4 -sign
has occurred in several of the previous texts.
However, at least two of the three other published exemplars of this text read a ka
sign, not a ke4-sign, at the end of line 9. This is hard to explain. There is no evidence for a
(phonological) change of le/ ) lal at this period; sporadic cases do occur in later S umerian,
but under different conditions.
One possibility is to see a long anticipatory genitive, of a kind not seen previously.
The essence of the sentence would be:
"Of Amar-Sin ... , to Enki his beloved lord" )
"To Enki, the beloved lord of Amar-Sin".
Although somewhat similar anticipatory genitives do occur in Sumerian, there seem to be
no exact parallels to this construction. Also, such an interpretation would not leave any
overt agent for the finite verb.
It is difficult to find a satisfactory explanation for the writing with the ka-sign. It is not
simply a scribal error, since the ke4 -sign seems to occur in only one exemplar; two
exemplars clearly have the ka-sign, and one exemplar is slightly damaged at the crucial
point.
The fact that the ke4 -sign is apparent in only one copy is in itself suspicious. It is
possible that this writing represents an (unconscious) attempt by the scribe to bring the text
into line with more common Sumerian morphology and orthography. (This assumes, in
fact, that the ke4-sign is indeed present, and that it is not an error on the part of the modern
day editor of the text; there is no photograph available of the text.)
 Lesson 14 163

-Titulature

Texts 1 3c and 14 use the title lugal-kalag-gf!, in place of nitag-kalag-g.!!. According to

Hallo, Amar-Sin was the first king of the Ur lIT Dynasty to adopt this title. He and his
successors used it "to the virtual exclusion of the older title".

-Terminology

Occasionally, S umerologists and Assyriologists will informally use an Akkadian

word, even when referring to its Sumerian counterpart. For example, even when
discussing the abzu, in a S umerian context, Sumerologists will not infrequently refer to "the
apsu". There is no theoretical or ideological reason for such practice; it reflects the fact that
Sumerologists learn Akkadian before they learn Sumerian, and also the fact that more is
usually known about the Akkadian word and its referent than about the Sumerian word.
164 Lesson 14

Text 1 4a
supplementary

This was inscribed on a cone.

 Lesson 1 5
This is another standard inscription of Amar-Sin. It may have been inscribed on the
pedestal of a statue.

Sign-list and voc abu lary

alam statue

mu name

ama mother

numun seed; offspring, progeny

sig9 (g)
1] to be narrow

>< kur to change

bU6 to tear out, to uproot

f>--- til to put an end to

1'- me

4t- Im

eb

Notes

alam Transliterated by some as alam, and by others as alan. Some Sumerologists believe
that alan is the older, and alam the later form. On the other hand, Lieberman reconstructs the
original form as /alag/; he believes that nasals in word-final position were neutralized with
respect to point of articulation. Others have postulated some kind of connection with the
word appearing in Akkadian as almu, in Hebrew as elem, "image".

barag-sig9 -ga The exact meaning is unsure. It may mean "narrow dais", barag.sig.a. It
is translated by the PSD as "socle (of a statue)". It was borrowed into Akkadian as
barasigQ, translated by the CAD as: "low socle for cultic purposes".

1 65
1 66 Lesson 15

kur This was equated with Akkadian aml and nakaru. Both have many meanings. For
nakaru in the D-stem, the CAD lists, among others, " ... 8. to discard an object (tablet, stela,
etc.), to remove an inscription ... 9. to clear away rubble, etc., to discard, remove from a
container, to demolish a building ... 1 1 . ... to place an object in a new location ... ".
Most frequently, kur governs a direct object (patient).

bU6 This is translated by the PSD as "to tear out", "to pull out", "to uproot", "to extirpate".
Although the meaning is clear, the precise reading is not. The sign is a ka-sign with an
inscribed kar-sign. This is most clearly seen in the Neo-Assyrian form of the sign: .
Sumerologists often transliterate such inscribed signs with an "x": KAxKAR.
Older works transliterate the ka-sign as bus (when necessary), KAxKAR as and
KAxSU as bu. Recent works, however, transliterate both KAxKAR and KAxSU as bu,
and add the fuller transliteration in parentheses: bu(KAxKAR); bu(KAxSU). However,
such a system can lead to confusion, because it is easy for the forms in parentheses to be
accidentally dropped in the mechanical process of printing. And, since the entire function
of transliteration is to provide a one-to-one correspondence of a specific cuneiform sign
with a specific transliteration, it seems counter-purposeful to use bu for two different signs.
Therefore, the older procedure is followed here, and the sign is transliterated as .
It is probable that the root of this word ended in some kind of Ir/-Auslaut; in Text 1 5
i t appears i n a verb form written i-bu 6 -re-. However, no Iburl value for this sign is re
cognized by the standard sign lists. A possible reading buzurs is recorded, but this may be
some other use. The problem deserves further study.
bU6 usually governs a direct object (patient).

til This sign has several meanings in Sumerian. In its reading as til, it is equated with
Akkadian gamaru, laqatu, and qatu. The CAD glosses qatu as " 1 . to come to an end, to be
used up, 2. to perish, 3. to become completed, finished, settled". In the causative stern,
uqtu is glossed as "to bring to an end".
 Lesson 15 167

Text 1 5

Column I Column I (continued)

* L:) f-*- EE
rf
M .ti
<H) A
$=I , :J

* mE1


lsWT> Fr1 1 The sign is partly restored from the

d plicate, No. 90039.

4
168 Lesson 15

Column II Column II (con tinued)

t=:1 :t .&-) r=CJ

.: :.: .
Fft2f:1< ;
T=f
.
<: tN. B-
.

*
-- .

x if\:f rt "
] rwti> " av=r
ii===> 1N n p
. ).
* } -<fPJ
I ,
T=:I m
* f-
 Lesson 15 1 69

Notes

11 9: The nam-sign and the U1!-sign are either partly effaced or poorly drawn on the
original. Not all scholars, in fact, believe that there even is a U1!-sign in this particular text.
But since this curse-formula occurs in several other texts, its restoration is relatively certain.

Transliteration Transcription
I 1: dAmar-dZuen Amarzuen
2: NibrukC1! Nibru.a
3: dEn-lil-Ie Enlil.e
4: mu-pad-da mu.pad.a
5: sag-us sagus
6: _ dEn-lil-ka e.Enlil.(a)k.a(k)
7: lugill-kalag-g1! lugal.kalaga
8: ' ki-ma
I !wI -Vnm
...! lugal.Vrim.a(k)
S
9: lugal-an-ub-da-lirnrnuz-ba-me lugal.anub.da.(k) limmu.bLa(k).me.(en)
10: alam-ba alam. bLa(k)
1 1 : dAmar-dZuen ---.! ki-ma Amarzuen kLaga.a. Vrim.a(k)
k' -1!g-!!TVnms
12: mu-bi-im mu.bLm
13: alam-ba alam. bLa(k)
II 1: lu ki-gy..b-ba-bi lu kLgub.a.bL0
2: i b-da-ab-kur-re-!! Lb.da.b.kur.e.0.a
3: barag-sig9-g1!-bi baragsiga. bL0
4: i-bU 6-re-!! L(b.)bur.e.0.a.(d)
5: dNanna Nanna
6: lugill-Vrimskt ma-ke4 lugal.Vrim.ak.e
7: dNin-gill Ningal
8: Ski-ma-
ama-Vnm k 4 ama. Vrim.ak.e
9: nam-U1!-ba-an-da-kur, ne namue.ba.n.da.kur.(e.e)ne.0
1 0: numun-na-ni numun.ani.0
1 1 : h-eb-til-Ie-ne ue.(L)b.til.e.ene.0
Translation
I 1: I am Amar-Sin,
2-4: proposed in Nippur by Enlil,
170 Lesson 15

5-6: patron of the temple of Enlil,

7 the mighty king,
S the king of Ur,
9: the king of the four quarters.
1 0, 1 2: The name of this statue is:
11: "Amar-Sin is the beloved of Ur".
11 5-6: May Nanna, the king of Ur,
7-8: and Ningal, the mother of Ur,
9: curse
1 3-2: the man who changes the place of this statue
3: and the man who tears down its pedestal,
1 0- 1 1 : and may they put an end to his offspring!
Commentary
1 . The first nine lines of this inscription are the same as in Text 13, except for the later
form of the royal title in line 7.
9. me is the first person singular enclitic copula, "I am". In older Sumerian, it is usually
written -me-en. The form of the second person singular is also -me-en. The third person
singular is -am. Thus, the paradigm for the singular of the enclitic copula is:

first person singular -me-(en)

second -me-@)
third -am (after a consonant)
-m (after a vowel)

The distribution of -me - -me-en does not exactly correspond to a difference in time.
In older Sumerian, -me-en is the norm, but in Ur Ill, both -me and -me-en occur, with -me
predominating. However, both also occur in later texts. It is not sure if this should be
regarded as an orthographic or as a phonological problem; it is discussed further below.
The regular form of the enclitic copula for the third person singular is laml, written
with the am-sign. After a vowel, the enclitic copula appears as Iml, as in line 1 0 of Text
1 5.
1 0. The next three lines give the actual name of the statue. The construction is an
anticipatory genitive, with an enclitic copula: "of this statue, its name is ... ". Thus, the bi of
mu-bi resumes alam: alam.bi.a(k) ... mu.bi.m.
1 1 . The actual name of the statue is: "Amar-Sin is the beloved of Ur". This is an
equational sentence, and so one might have expected to find an enclitic copula. However, it
is not uncommon to find simple equational sentences without a copula, and this name may
be such an instance. It is also possible that the name is not a complete sentence, but rather
is a noun phrase with an appositive: "Amar-Sin, the beloved of Ur"; this is discussed
 Lesson 15 171

further below.
The next few lines are rather complicated. Lines I: 1 3 through IT: 4 are all the
comitative complement of the verb in IT: 9. This complement includes a relative marker iliD
governing two relative clauses. The first is marked by the - at the end of line IT: 2. The
second is marked by the -ft at the end of line IT: 4. All of this is embedded inside an
anticipatory genitive:
"of this statue, the man who changes its position and tears down its pedestal" )
"the man who changes the position of this statue and tears down its pedestal".
The anticipatory genitive in I: 13 is resumed by the -bi in IT: 1 and 11: 3.
11 1 . ki-gub-ba = ki.gub.a, "standing place" or something similar. Presumably, gub-ba
-

is a passive participle in .a.

2. Since the sense being conveyed is future, the verb is put into the maru. kur is a member
of the affixation class, so forms its maru with the maru-suffix .e, hence kur.{;, written kur
re.
The initial lil of the ib-sign represents the conjugation-prefix i.
There are two Ibis in this particular prefix chain; it is easier to look at the second Ibl
first. Since this is a maru form of the verb, the Ibl in the slot immediately preceding the
verbal root cross-references the direct object , kiguba.bi.0.
The da is the dimensional-prefix which cross-references a nominal phrase in the
comitative case. The Ibl before the da is an element not yet seen. Before the dimensional
prefixes .da (cross-referencing the comitative .da), .i (cross-referencing the terminative
.e), and .ta (cross-referencing the ablative), it is possible for an "optional pronominal
prefix" to appear. For the third person, these prefixes are: .n for the animate, and .b for the
inanimate (forms for first and second person are discussed below). These prefixes help
cross-reference the nominal phrases occuring in the sentence; they do not convey any new
information. Here, the Ibl refers back to alam of line 10. The use of these pronominal
prefixes appears to be purely optional (at least, no-one has figured out any rules for their
distribution); they did not appear in any of the previous texts used in this book, and only
show up sporadically in the remaining texts. This is not apparently a problem in
orthography. That is, unlike the personal-affixes appearing immediately before the root,
one should not assume that these pronominal-prefixes were always present.
Thus, .bda represents the comitative dimensional-prefix, with an optional pronominal
prefix.
An obvious problem here is that there is no comitative nominal phrase in the sentence
for the dimensional-prefix to cross-reference. In fact, in this verbal prefix chain there oc
curs both a dimensional-prefix (.bda) and a personal-affix (.b). However, there is only one
noun phrase that these could cross-reference, ki-g!!h-ba-bi. Furthermore, this particular
verb seems to normally be construed with a direct object. Therefore, the personal-affix .b
cross-references the direct object, and the dimensional-prefix .bda does not refer back to
any particular nominal phrase.
As was mentioned in Lesson 12, there are numerous cases where a comitative
dimensional-prefix appears in the verbal prefix chain, but with no corresponding comitative
nominal phrase in the sentence. There is probably not just one single rule governing th(
172 Lesson 15

appearance or non-appearance of the comitative dimensional-prefix. More likely, there are

several different factors at work, which have not yet been unraveled.
To summarize the verb form:

i . b . da b . kur . e . 0 a
(1) (2) (3) (4) (5) (6) (7) (8)
(1) conjugation-prefix
(2) optional pronominal-prefix
(3) comitative dimensional-prefix
(4) personal-affix cross-referencing patient
(5) verbal root
(6) maru-suffix
(7) personal-affix cross-referencing agent
(8) nominalizer

The entire clause, nominalized in -!!, stands in apposition to ill, forming a relative
clause.
4. i-bu6-re-!! = L(b).bur.e.a. i is the conjugation-prefix. However, the verb has no dimen
sional-prefix, nor (on the surface) does it have any personal-affix cross-referencing the
apparent direct object, baragsiga.bi.0. Since the verb form is in the maru, one might have
expected to find /b/ immediately preceding the verbal root. There are at least two possible
reasons for its apparent absence. Lines 3-4 are roughly parallel in form and content to lines
1 -2. It is possible that its use in the verb form in line 4 would have been felt as redundant;
that is, the parallelism in construction permitted deletion of the personal-affix .b. Some
scholars have suggested that personal-affixes are only present (and therefore only written)
in ambiguous contexts.
However, there may be an entirely different reason for its absence. Perhaps there was
a phonetic reduction of /ibburea/ ) /iburea/. That is, the problem may be phonological, not
morphological. As is often the case in Sumerian, it can be difficult to determine whether a
problem is orthographic, phonological, or morphological in nature.
4. The assumed comitative case-marker .(d) at the end of this line marks lines I: 13
through IT: 4 a s the comitative complement o f the verb nam. . .kurS i n line 11: 9. The same
use of the assumed comitative with this verb occurred in Text 1 2.
6-8. The two agents of the verb forms in lines 9 and 1 1 are both marked by the ergative
case-marker .e.
9 . .nda is the comitative dimensional-prefix plus the optional pronominal-prefix .n. Here
the animate form is used, because it refers back (essentially) to "the man who ... ".
All the verb forms seen up to this point have been singular. Here, there is a plural
agent ("Nanna and Ningal"), and so the verb must be put into the plural (Sumerian has no
dual). The plural third person of a maru verb is usually written with a suffixed --ne.
S umerologists have different understandings about the morphology implied by this writing.
One analysis sees this writing as reflecting: e.ene.0. The .e is the mat-u-suffix; .ene is the
plural marker; and .0 is the personal-affix cross-referencing the agent. Another analysis is
 Lesson 15 173

to isolate the morphemes of this ending as: e.0.ene. A third analysis is to read the ne-sign
as de, reflecting quite a different understanding of the morphology. This is a very difficult
issue to resolve. Here, the first analysis has been followed.
A problem in this particular verbal form is the fact that only -ne is written, not --ne.
How should this writing be understood?
The Falkenstein school would read the first sign with an overhanging vowel, that is,
kure2' A variant of this solution is to read the kur-sign as kUfUS : kUfU s-ne would represent
kur.e.ene.0. The second lul in the sign kUfUS would represent the assimilation of the un
derlying lel to the first lu/. There are numerous other cases where the marii-suffix .e is
assimilated to an I ul-vowel of a verbal root.
But has been stated several times, the mnemonic nature of the script may have meant
that there was no need for the scribes to write down the full forms of the morphemes.
To summarize the verb form:

nam be ba n da kur . e . ene 0

(1) (2) (3) (4) (5) (6) (7) (8) (9)
(1) nominal component of compound verb
(2) modal-prefix
(3) conjugation-prefix
(4) optional pronominal-prefix
(5) comi tative dimensional-prefix
(6) verbal root
(7) mara-suffix
(8) pI ural marker
(9) personal-affix cross-referencing agent

This line is essentially the same curse-formula which occurred in Text 1 2. It is

instructive to compare the verb forms of the two texts:
Text 12 nam-bf!-ba-da-kurs-
Text 1 5 nam-bf!-ba-an-da-kurs-ne
The first difference is the presence in Text 15 of the animate optional pronominal
prefix .n before the comitative dimensional-prefix. It is precisely cases such as this - two
texts with a minimum of variation - which seem to show that such pronominal prefixes are
indeed optional.
The second difference is in the number of the verb. In Text 12, the agent of the verb is
singular ("Gilgamesh"); in Text 1 5, the agent of the verb is plural ("Nanna and Ningal").
1 1 . b is the regular form of the desiderative modal-prefix. In its two previous
occurrences, Ibe/ ) Ibal before the conjugation-prefix Iba/.
The next slot in the verbal prefix chain should be the obligatory conjugation-prefix. In
this particular case, the conjugation-prefix 1 has contracted into the I el of the modal-prefix .
.b cross-references the direct object, numun.ani.0. Since numun refers to "future
descendants, progeny", it might seem a little surprising to see numun.ani.0 cross
referenced by .b (normally used for inanimates) instead of by .n (normally used for
animates). However, .b is frequently used for what might be considered "collectives" (both
174 Lesson 15

of animates and inanimates), and here numun was probably felt as a collective.
Since til is a transitive verb, b is used with the maru form of the root. til is a member
of the affixation class, so its maru-root is ti1.e. The verb is written as expected, til-Ie-ne.
To summarize the verb form:

be i b til e ene o
(1) (2) (3) (4) (5) (6) (7)
(1) modal-prefix
(2) conjugation-prefix
(3) personal-affix cross-referencing patient
(4) verbal root
(5) maru-suffix
(6) plural marker
(7) personal-affix cross-referencing agent

Discussion: structure
This text is composed of three sentences:

I: 1-9 nominal sentence

I: 10- 1 2 nominal sentence
I: 1 3-11: 1 1 verbal sentence

The structure of the verbal sentence, in essence, is:

I: 13-II: 4 accompaniment
11: 5-6 agent 1
11: 7-8 agent2
11: 9 verb
11: 1 0 patient
11: 1 1 verb

In a famous monograph entitled Das appositionell bestimmte Pronomen (1932), Poe

bel tried to show that the copula in Semitic and in Sumerian could also be used to express
apposition. In Text 15, the translation of the beginning of the inscription would be: "I,
Amar-Sin, the one nominated ... ". That is, the first nine lines would not form a complete
sentence, but rather would form a kind of casus pendens or fronting for emphasis. Simi
larly, the name of the statue in line 1 1 may be a noun and appositive, and not an equational
sentence.

- Sign formation

In Lesson 3, it was pointed out that some cuneiform signs are in origin combinations
of a pictographic sign with an inscribed phonetic indicator of some kind. The ama-sign is
thought to be one of these. It has an inscribed digir-sign, one of whose phonetic readings
is am6 . Therefore, it has been suggested that this component of the sign is a clue to the
 Lesson 15 175

pronunciation. Unfortunately, it is difficult to make this square with the fact that the rest of
the sign appears to be the pisan-sign, which basically means "box" of some kind. It has
also been proposed that there is some obscure symbolism involved, "mother" being
represented as a "divinity" within a "box"!

- Loan words

In Text 6, the word barag occurred; this term was borrowed into Akkadian as parakku.
The amissable /g/-Auslaut shows up in Akkadian as a geminated voiceless consonant.
barag-sig9-gi!., however, was borrowed into Akkadian as barasigu. The amissable / g/
Auslaut of the barag element, here in syllable-final position, does not appear. However, the
intervocalic / g/ of gg-gi!. (sig.a) remains. The word-initial /b/ of barag is treated dif
ferently in each loan word. In parakku it is reflected as /p/, but in barasigfi it is reflected as
/b/. This shows that barasigu is a later borrowing into Akkadian than parakku. It is us
ually assumed that in relatively older loan words from S umerian, Sumerian voiced stops are
reflected in Akkadian as voiceless stops. In relatively later loan words, the same Sumerian
voiced stops are reflected as voiced stops. Needless to say, not enough is known about the
phonetics of either Sumerian or Akkadian to explain exactly what has happened. However,
such differences are one way that the entry of loan words into another language can be
dated relative to each other.

- Conjugation -prefixes

In the model of the Sumerian prefix chain used throughout this book, the use of the
conjugation-prefix is obligatory; a conjugation-prefix is present in every finite verbal form.
Therefore, in line 11: 1 of Text 1 5, it is assumed that a conjugation-prefix 1 has assimilated
into the modal-prefix b. Similarly, in Text 22, this view assumes the presence of a
conjugation-prefix 1 after the cohortative modal-prefix g.!!, although the verb form in Text
22 is written g.!!-an-!!-il.
Although both the desiderative modal-prefix b and the cohortative modal-prefix g.!!
are frequently followed by such conjugation-prefixes as mu and ba, there appear to be no
instances of writings of the type *b-l or *g.!!-l. If the 1 is in fact present, one might expect
to find at least a few occurrences of it being written (to judge by similar phenomena in the
script). However, none apparently occur. This means that the assumption that the attested
spellings all represent assimilation may not, in fact, be correct.
Jacobsen, for example, believes that the conjugation-prefix 1 is "incompatible" with the
cohortative modal-prefix g.!!. That is, the semantic information conveyed by 1 does not
permit it to co-occur with g.!!. This may mean that at times the modal-prefix g.!! is followed
by no conjugation-prefix. The up-shot is that not every finite verbal form contains a
conjugation-prefix.
This problem cannot be resolved here. However, it should be kept in mind that the
general principle stated in this book - that conjugation-prefixes are obligatory - may need
modification, particularly in the case of certain modal-prefixes.
176 Lesson 15

- Pronominal-prefixes

The use of pronominal-prefixes before certain dimensional-prefixes is not uncommon.

In the singular, the basic fonns of these prefixes appear to be:

first person singular .0

second .e
third animate .n
inanimate .b

The form of the first person singular was probably not simply .0. Some irregular
writings have led scholars to speculate that the first person singular marker had either a
vocalic component (perhaps lel, similar to that of the second person), or a consonantal
component (perhaps a semi-vowel or glottal consonant). The fonns of the plural are much
less clear.
If more than one dimensional-prefix occurs in a verbal prefix-chain, it appears that
only the first dimensional-prefix can have an optional pronominal-prefix.
An older view of these pronominal-prefixes was that they were obligatory, and that
their relative infrequency in texts is due to vagaries of orthography. In this view, for
example, every .da dimensional-prefix which cross-references a third person should be un
derstood as: (n.)da. Thus, the verb fonn in Text 1 2, nam-b-ba-da-kur5 -, should be un
derstood as: nam.be.ba.(n.)da.kur.e.0, instead of: nam.he.ba.da.kur.e.0. However, this
view is not as widely accepted today, because the actual number of cases where an .n or .b
appears before an appropriate dimensional-prefix is much less than the number of cases
where they do not appear.

-Dimensional-prefixes

In line 2, the verb kur was used with the dimensional-prefix -da, although the sentence
contains no nominal phrase in the comitative case. In some cases, verbs have become
lexicalized with certain dimensional-prefixes; that is, the verb will frequently (sometimes
always) have a certain dimensional-prefix, even if no corresponding nominal phrase occurs.
Occasionally a noun phrase will be marked with a certain case-ending, but the
dimensional-prefix used in the corresponding verbal phrase will be different than e,xpected.
For example, a nominal phrase in -e (the tenninative) may be resumed by the dimensional
prefix nonnally used for the locative (-ni-). In some cases, this may have happened
because of a historical change in the rection of a noun phrase. That is, at one time a noun
phrase may have been marked by one particular case, but in time the case which was used
changed. However, the dimensional-prefix, being more closely bound, did not change.
This can lead to such cases as a tenninative being cross-referenced by a locative. In the
case of kur, for example, its complement may originally have been in the comitative case (or
perhaps the ablative case), but it eventually shifted to the absolute case. However, it carried
along its dimensional-prefix, resulting in such cases as line 11: 2, where the verbal prefix
chain has both a dimensional-prefix and a personal-affix, yet there is only one nominal
 Lesson 15 177

phrase for both of them to govern.

Similarly, the rection of the first element of a compound verb may change. In earlier
texts, it may be in one of the adverbial cases; in later texts, it may be construed as a patient.
Sometimes this change may be due to Akkadian influence. Unfortunately, there is rarely
enough data to prove such hypotheses for any particular verb.
The analysis of 11: 1 -2 presented here is not the only one possible. 11: 1 was explained
as the patient of the verb, and the da in the verbal prefix chain as a frozen use of the
comitative dimensional-prefix. However, it is known that the verb kur sometimes takes its
complement in the ablative case. Therefore, it has been proposed to understand line 1 as:
kLgub.a.bi.(ta), with the ablative case-marker not expressed in writing. The dimensional
prefix which cross-references the ablative is Ita/, normally written ta. Therefore, the verb
form in line 2 would have to be understood as: Lb.ta.b.kur.e.0.a. The fact that the text
clearly shows a da-sign, and not a ta-sign, would seem to mitigate against such an
interpretation. However, other cases of the ablative dimensional-prefix appearing on the
surface as da instead of ta are known; the reasons why are unclear. In our particular text,
there could have been assimilation of voice: Ibta/ ) Ibda/. However, other such cases of
an apparent da for ta are less amenable to phonetic explanations of this sort.

-Enclitic copula

The fact that the first and second persons of the enclitic copula occur both as -me and -
me-en admits of several possible interpretations. The problem may have been orthographic:
word and syllable-final nasals often seem not to be written in S umerian (cf. mu-na-du for
mu.na.(n.)du). It may reflect the phonetic process of Sumerian dropping word-final nasals;
in this case, the forms written with -en are to be regarded simply as morphographemic or
archaic writings. It has also been posited that the writings -me and -me-en represent one
pronunciation, /me/. The original word-final nasal was lost, producing a nasalized vowel,
which the script could not well represent.

- Proper names

Votive objects of all kinds were often given names. This has been studied by Gelb,
who points out that such names are often complicated sentences in their own right. For
example, a statue of Gudea dedicated to Ningirsu was given the name (Gelb's translation):
"Ningirsu, -the-king(=god)-whose-heavy-might-the-world-cannot-bear, -has-decided-good
destiny-for-Gudea, -who-built-this-temple" (1956:66).

- History

In addition to the copy of the text reproduced above, a late, Neo-Babylonian copy from
the seventh century BC is also preserved. It was inscribed on what was apparently a model
pedestal. This copy is interesting because of the presence of several errors in the Sumerian.
Also, it has a colophon written in Akkadian, which seems to say that the model was to be
used in an "exhibition" (tamartu) of some kind. The page following is a photograph of two
sides of this Neo-Babylonian version of the inscription.
The following quotation from Woolley describes this object, and also says something
178 Lesson 15

about the tenor of the times which produced it:

A little way apart lay a small drum-shaped clay object on which were four
columns of writing; the first three columns were in the old Sumerian
language, and the contents of one at least were familiar to us, for we had
found it on bricks of Amar-Sin, king of Ur 2046-2038 B.C., and the other
two were fairly similar; the fourth column was in the late Semitic speech.
"These", it said, "are copies from bricks found in the ruins of Ur, the work
of Amar-Sin king of Ur, which while searching for the ground-plan [of the
temple] the Governor of Ur found, and I saw and wrote out for the marvel of
beholders". The scribe, alas!, was not so learned as he wished to appear, for
his copies are so full of blunders as to be almost unintelligible, but he had
doubtless done his best, and he certainly had given us the explanation we
wanted. The room was a museum of local antiquities maintained by the
princess Ennigaldi-Nanna (who in this took after her father, a keen
antiquarian), and in the collection was this clay drum, the earliest museum
label known, drawn up a hundred years before and kept, presumably
together with the original bricks, as a record of the first scientific excavations
at Ur. ... We shall see further examples of the archaeological spirit that
prevailed in the latter days of Babylon, but undoubtedly it was reinforced by
a pathetic superstition that looked back across the uncounted ages to the
fabulous beginnings of things when men and gods were scarcely to be
distinguished and "there were giants in the land in those days" (1982:252,
23 1).
Lesson 15
179
 Lesson 1 6
This is another door socket of Amar-Sin.

Sign-list and vocabulary

$ hH honey

JP i oil, fat

' nun prince, noble

rv i-nun butter

.
vme, wme
.

siskur2 sacrifice

Wig to cease

!!. and

nu not

Notes

11\1 It is thought that the bee is native to Syria, but not to Mesopotamia. So Iftl (and its
Akkadian equivalent dgpu) may be a Syrian import. It has also been speculated that hU and
dispu were actually a kind of syrup made from fruits.

i-nun Literally, "oil of the prince". Some scholars interpret this as "butter", others as
"ghee" (that is, clarified butter). The latest discussion (K. Butz, 1973-74:37) interprets it as
"Butterschmalz" ("butterfat"). The Akkadian equivalent of i-nun is bimetu; this is trans
lated by the CAD as "ghee".
On the surface, i-nun would appear to be a noun-noun compound, of the type dis
cussed in Lesson 9. However, there are spellings in other texts which show that this is a
genitive phrase, "oil of the prince", Lnun.a(k). For example, in Gudea this expression is
written both i-nun and i-nun-na. The term i-dug-nun-na, = i.dug.nun.a(k), "good oil of the
prince", also occurs, but it is not sure what this means.
In older works, the word for "oil, fat" is transliterated as hi, instead of i.

gdtin The word for "wine" also appears as tin. gestin, therefore, may originally have

181
1 82 Lesson 16

meant "wine-tree, vine, grape", etc.; the meaning then became extended to "wine". The
Akkadian equivalent, karanu, is glossed by the CAD as: " 1 . wine, 2. grapewine, 3.
grapes".

siskur2 This word is normally written just as it appears here, by graphic reduplication of
one sign. This sign appears to be the amar-sign inside of which is the barley-sign, Se.
Thus, the original, pictorial significance of the sign may have been "grain-fed cattle", or
something similar.
Sometimes the word Isiskurl is written with only one, instead of two, signs. In this
case it is transliterated as siskur. Earlier Sumerologists thought that the writing with one
sign was a singular, and the writing with two signs was a plural; this means that the writing
with two signs was sometimes transliterated as siskur-siskur. The word is further dis
cussed in Lesson 19.
The pronunciation of the sibilants is unsure, and so it is also transliterated as siskur2 '
sizkur2' Siskur2 , etc.

ilig Or, silig. This root is very uncommon in verbal forms; it is mostly used in participial
or infinitival constructions.
 Lesson 16
1 83
Tex t 1 6
184 Lesson 16

Transliteration Transcription
1: dEn-lil Enlil
2: lugill-kur-kur-ra luga1.kur.kur.a(k)
3: lugill-ki-.M-M-ni-ir luga1.kL aga.a.anLr
4: dAmar-dZuen Amarzuen
5: dEn-lil-Ie Enli1.e
6: NibrukC Nibru.a
7: mu-pad-da mu.pad.a
8: sag-us sagus
9: _dEn-lil-ka e.Enli1.(a)k.a(k)
1 0: lugal-kalag-g luga1.kalaga
I !g-
1 1 : ...! ' ki-ma
I Unms luga1.Urim.a(k)
12: lugill -an-ub-da-limmurba-ke4 luga1.anub.da.(k) limmu.bLak.e
13: lal i-nun e lal inun
14: y getin u getin
15: ki-siskurTra-ka-na kLsiskur.ak.anLa
16: nu-i1ig-g nu.i1ig.e(d).0.0
17: mu-na-an-du mu.na.n.du.0
Translation
1: For Enlil,
2: king of the lands,
3: his beloved king -
4: Amar-Sin,
5: proposed by Enlil in Nippur,
8: patron of the temple of Enlil,
10: the mighty king,
1 1 : the king of Ur,
1 2: the king of the four quarters -
17: built -
13: the temple (where) honey, butter
14: and wine
15: in his place of sacrifice
 Lesson 16 185

16: shall not cease.

Commentary
12. The .e marks the agent, which spans lines 4 through 12.
1 3 - 1 6. These lines are the direct object (patient) of the verb in line 17; they are marked as
such by the final .0 at the end of line 16.
means "temple"; the rest of the text is apparently a relative clause modifying . Up to
this point, all the relative clauses which have occurred have been verbal sentences. Here,
however, the relative clause is a nominal, equational sentence: "honey, butter and wine in
his place of sacrifice are a non-ceasing thing". The syntax of nominal relative clauses is
little understood. They consist of a nominal element as subject and a nominal element as
predicate; the enclitic copula is not used. Unlike the case with verbal relative clauses, no
nominalizer is used. Also, no relative marker is used. This entire equational sentence is in
apposition to the noun .
Although Sumerian does not use any relative marker, English needs a relative adverb
of some kind, such as "where".
14. Sumerian does not normally use any conjunction between nouns; instead, it prefers to
conjoin them directly. However, Sumerian also has an option of using the conjunction 1l
between nouns or nominal phrases (rarely between verbs). Here, it is used between the
second and third nouns of a list: "honey, butter and wine".
1l is not a native Sumerian word; it is borrowed from Akkadian. It is not common for
languages to borrow such syntactic devices as conjunctions. This is an indication of the
Sumerian language giving up some ground before the onslaught of Akkadian. 1l turns up
as early as the Tell Abu Salabikh tablets (2600 BC). According to the editor of these
tablets, Robert Biggs, "even at this early date Sumerian may have been under a heavy
Semitic influence" (1974:32). (Post Ur III texts also occasionally use the Akkadian
conjunction -ma.)
1 5 . The final .a is the locative case-marker. It is not resumed by any dimensional-prefix.
1 6. nu is the general, all-purpose negative marker in Sumerian. It can appear before verbal
forms and before nominal forms.
When used with verbal forms, nu falls into the category of modal-prefix. Thus, it
never co-occurs with any other modal-prefix. It is followed instead by a conjugation
prefix. (The conjugation-prefix 1 often assimilates into the lul of nu.)
ilig is apparently an active participle from the verbal root "to cease" .
.ed is an element not seen before. It is one of the most puzzling of all Sumerian
morphemes. It normally follows manl-forms of the verb; its use with bam!!!-forms is very
rare. It is also used with both active and passive participles. With active participles, it
occurs as: du.ed.0. With passive participles (much rarer), it occurs as: du.ed.a. The form
in Text 16 is: ilig.ed.0.
The meaning of .ed has been much discussed. It must have something to do with the
"future", although the exact nuance being conveyed is unsure. In its very rare occurrences
with the barn!!!, it appears to have a future-perfect sense. According to Jeremy Black,
Its reference seems to be to future events, although its use in descriptive
1 86 Lesson 16

passages suggests a connotation of vividness. More important, it marks an

action as not yet begun at the moment of observa tion (which can be a
moment in the past) (1984: 1 1 8).
According to Jacobsen, .ed is the
mark of pre-actional aspect indicating prospectiveness of the action as
present at the point in time the speaker has in mind. Attention is thus not on
the action as future but on its prospectiveness as present (1965:267).
Because Idl is an amissable consonant, it often does not show up in the script. This
can make it very difficult to recognize.
Line 16 forms the predicate of the nominal relative sentence. This predicate is a
nominal form: "something which will not cease in the future", i1ig.ed.0. Thus, the first
.0 in transcription marks ilig.ed as an active participle. The second .0 marks the entire
complex as the patient of the verb in line 17.
At first glance, it might seem possible to interpret nu-ilig-g as a verbal, not a nominal
form. However, this analysis would seem to be precluded, because of the lack of a final -.!!
in line 16. If it were a verbal form, it would need to be nominalized in -..!! to serve as a
relative clause.
17. mu-na-an-du mu.na.n.du. This is the first time in this body of texts where the ani
=

mate personal-affix .n, cross-referencing the agent in the bamn!, appears explicitly in the
script.
There is no dimensional-prefix cross-referencing the locative nominal phrase in line
15.
Discussion: structure
The structure of this text is:
[Enlil, lugal.kur.kur.a(k), lugal.kLaga.a.ani].r benefactive
[Amarzuen, Enlil.e Nibru.a mu.pad.a, agent
sagus.e.Enlil.(a)k.a(k),lugal.kalaga,
lugal. Urim.a(k),
I ugal.anu b.da. (k)limmu.bLak].e
[e lal inun u getin kLsiskur.ak.anLa patient
nu.ilig.e(d).0].0
mu.na.n.du.0 verb
Relative clauses consisting of a nominal sentence, such as lines 1 3-16, are not
common in Sumerian, and so there is some question about the analysis. It is possible that
these lines should be understood as the name of the temple: "Amar-Sin built a temple
(whose name is) 'honey, butter and wine in his place of sacrifice shall not cease"' . In such
constructions, Sumerian does not always use the word for "name" (mu); instead, it conjoins
the name directly, as an appositive. An example occurs in Text 19: bad-Mar-tu Mu-ri-iq
Ti-id-ni-im, "the Martu-wall (whose name is) 'Muriq-Tidnim" ' .
 Lesson 16 187

- .ed
There is no standard term used to refer to .ed. It can appear both on finite verbal
forms, and on nominal forms derived from verbal roots. On finite verbal forms, it occurs
with both marii-roots and bamt!!-roots (although there has been much argument about this).
It seems to occur with both transitive and intransitive-passive forms. On nominal forms, it
occurs on the "infinitive", the active participle, and the passive participle.
Because the Idl is amissable, and because the lel can assimilate into other vowels, it
is not always easy to determine even if it is present or not. The assumed Id! does not show
up in Text 1 6, and it is therefore also possible that some other interpretation of line 1 6 is to
be preferred. In spite of several recent attempts to unlock its morphology and syntax (the
most recent is that of Gerd Steiner [1980]), it remains elusive.
-Participles
The traditional view of Sumerian is that it has two participles, an active and a passive
intransitive (Poebel used the terms nomen il,gentis and nomen actionis). The active ends in
.0, and the passive-intransitive in .a. Each participle can also appear with the element .ed.
This yields four possibilities:
X.0 X.a
X.ed.0 X.ed.a
However, a fair number of exceptions seem to occur. There are cases of participles in
.0 which seem to have a passive meaning, and participles in .a which seem to have an
active meaning. For example, the form in Text 1 6 is mig-g, presumably for ilig.ed.0.
Here it is being used intransitively, and therefore one might have expected a participle in .a.
The situation was undoubtedly more complex than the above table would indicate. Most
discussion of non-verbal forms in Sumerian has applied Indo-European or Semitic
grammatical categories and terms to the S umerian forms.
The most recent, full-scale, study of the participles is that of I.T. Kaneva (1970). She
analyzes the forms differently:
X.0 "transitive participle of the imperfect aspect"
X.a "transitive participle of the perfective aspect"
X.a "intransitive participle"
However, she does not apparently recognize the existence of reduced relative clauses,
whose existence complicates the picture.
- Personal-affixes
This text marks the first time that the personal-affix .n cross-referencing the bamt!!
agent appears in the writing in the body of texts studied here. As discussed in Lesson 14,
this is part of the on-going process of the Sumerian writing system becoming more and
more explicit.
188 Lesson 16

- Conjunctions
The native Sumerian conjunction meaning "and", linking nouns and nominal phrases,
is -bi-da, suffixed to the second noun: an-ki-bi-da, "heaven and earth". It is thought that bi
here is the possessive-suffix, and da is the comitative case-marker. Thus, this originally
meant "the heavens along with its earth".
Sumerian also has a conjunction meaning "and", linking verbs: -in-g, occurring on
the last verb of a series of two or more, in the position after the modal-prefixes but before
the conjugation-prefixes. No examples seem to occur in the Ur III royal inscriptions. A
late example is: ... za- in-g--zu, "you also know", for: inga.i.e.zu. Because in-g is
used so infrequently, it may have had some emphatic value, instead of straight co
ordination.
- History
This door socket (along with another, having a different inscription) was found in
what has been described as the "small shrine" in Nippur, part of a large temple complex
dedicated to Enlil. The door socket bears traces of an inscription of Lugal-kiginne-dudu,
who ruled in Uruk approximately 2400 BC (that is, some three hundred years before
Amar-Sin); this early inscription was also dedicated to Enlil. Thus, Amar-Sin's builders
did not fashion a new door socket - rather, they engraved their inscription upon this
already-extant door socket, which presumably they had uncovered in their work of
rebuilding the Enlil temple.
-Literary parallels
The combination "honey, butter and wine" occurs in other Surnerian texts. Sometimes
the three elements are listed in the same order as in this inscription; sometimes the order
varies. For example, in a literary text entitled "Nanna-Sin's Journey to Nippur", there
occurs: "May Nanna-Sin make butter, honey, and wine (i-nun lal geStin) abundant"; this
particular line is repeated verbatim five times.
The expression i-nun has recently been found in one of the Sumerian-Eblaite voca
bulary lists. Unfortunately, only the Sumerian for this particular expression is given, not its
Eblaite equivalent. One of the other entries in this vocabulary list, however, is for "good
oil". The Sumerian column is i-dug; the Eblaite equivalent is sa-rna-nu t-bu.
 Lesson 16 189

Text 1 6a
supplementary

This inscription was engraved on a small bead of agate, used as a votive object.
Because of its small size and somewhat irregular surface, the signs are less than elegant.
No photograph is available.

..... .. ' \'

'f Y \. // (
,-It ',,;1:1 j ;;;< ->/
1 90 Lesson 16

Notes

Commentary

8. .!l serves as a conjunction between two nominal phrases.

9. Nin-tur-dumu-g!!l O. This appears to be a personal name, although there do not seem to
be any parallels to such a formation ("Nintur is my child" ?).
1 0. Both here and in line 7, the nominal phrases are marked by the terminative case-marker
-e.
1 6.It may be useful to compare the different forms of the standard curse-formula which
have occurred:
Text 12 nam-b-ba-da-kur5-
1 6a nam-b-ba-da-kur5"re6
15 nam-b-ba-an-da-kur5"ne (plural)
The difference between Text 12 and Text 16a is in the orthography of the maru-suffix.
In Text 1 2, a morphemic spelling is used. In Text 1 6a, the spelling follows the more usual
rules of Sumerian orthography.
 Lesson 1 7
This is a large door socket of Amar-Sin.

\9\
1 92 Lesson 1 7

Sign-list and vocabulary

J>Tf Kar-zid-da Karzida (GN)

f kar quay, pier; market-place

o utu sun

kalam land

<crr g-Illir giparu (part oftemple complex)

en priest, priestess

zid W) effective, true

<:=!> ul remote, distant

( lrng Gill) bright, pure, holy

. kur9 (ku4) to enter

>{>- sud4 13) to be long

JP li

tt>T ta

un

o:J ab

Notes

Kar-zid-da The name means "the effective quay", kar.zid.a. Karzida was apparently a
quay at Gaesh (written Ga-eSki and Ga-eSs ki). The latter was close to Ur, and evidently
had some kind of cultic connection with it; it had its own Nanna-temple. Neither the site of
Karzida nor of Gaesh has been identified.
kalam This word is apparently only used to refer to Sumer itself. It only occurs in the
singular, and is especially common with a possessive-suffix. The original pictographic
value of the sign is unknown. This word is further discussed in Lesson 19.
gi6 -pAr This was a section of the temple complex; it was where the en-priest or en-
 Lesson 1 7 193

priestess lived. It was borrowed into Akkadian as giparu. It is discussed further below.
en Besides its use in political contexts (where the conventional translation is "lord"), en
can also be used to refer to a specific kind of priest or priestess.
zid It is difficult to pin down a precise meaning for zid; it can mean something like
"someone (or something) that does what he/it should be doing". The Akkadian equivalent
is kInu, glossed by the CAD as: " 1 . true, reliable, just, 2. honest, decent, loyal, 3. correct,
normal, regular, sound, legitimate".
ul This sometimes appears as ul, as in, for instance, uru-ul, "the primeval city". However,
sometimes an lit follows, resulting in graphic reduplication: ul-H . And sometimes ul or
ul-li is followed by la/, presumably the same nominalizer seen in such adjectives as kalag
g. In Text 17, it is written: uHi-.
It is probably related to ul meaning "bud" of a flower.
kur9 In older works, usually transliterated as tu; some modern-day Sumerologists still
prefer this reading.
sud4 The verbal root meaning "to be long" can be written in two different ways in
Sumerian. The most common writing is: )---- , read as sud (or su). However, it can also
be written (as in Text 17) as: >t>- . Older works tended to understand this last sign as
gid, but most scholars now read it as sud4 (or su 1 3 )'
It can be seen that the sud4 -sign is basically just a non-gumlfied version of the sud
sign. In general, it is always possible to use a non-gunufied sign in place of a gunufied
sign. For example, the gi-sign is frequently found in place of an expected g4. In Lesson
1 , it was mentioned that the name of the city of Ur is usually spelled Se-ab, but also Se
unug. It is possible that the unug-sign is a gunufied ab-sign.
This is why >{>-- is now read sud4 instead of gid. At one time, it was thought that
Igidl and Isud! were two different roots for "to be long", but now it is assumed that they
are just variant writings of the same root I sud!.
The final consonant of the root was probably the Idr/-phoneme discussed in Text 12.
In Text 17 it is followed by the re6 -sign (with its initial Ir/).
 Lesson 1 7
1 94
Te xt 1 7
 Lesson 1 7 1 95

Notes

1 1, 18. In this text, the kalam-sign and the un-sign are very similar in shape. The histo
rical relationship between the two signs is complex. It is usually assumed that in origin
they were formally distinct. In carefully executed texts, they are still distinguished through
the Old Babylonian period. For example, in the Code of Hammurapi, the kalam-sign ap
pears as: , and the un-sign appears as: Iffif- -t .
However, the un-sign (more properly, yg; its value was either lugl or lugu/) means
"people", and there is a close semantic connection between "people" and "land". This
might mean that the signs were not formally distinct in origin, but perhaps became
secondarily differentiated.
By the Neo-Assyrian period, the two signs fall together into one sign. There are
several Neo-Assyrian signs which represent the conflation of two or more signs which in
origin were different signs. For example, the Neo-Assyrian ku-sign is a continuation of
several different box-shaped signs, which have all fallen together in shape.
17. The nu-sign is poorly drawn; a better drawn version appears in line 1 8. There appears
to be an extra horizontal stroke, running into the top of the du-sign.
Transliteration Transcription

1: dNanna Nanna
2: Kar-zid-da Karzida.(k)
3: lugal-ki-g-g-ni-!r lugal.kLaga.a.ani.r
4: dAmar-dZuen Amarzuen
5: dEn-lil-Ie Enlil.e
6: NibrukCi! Nibru.a
7: mu-pad-da mu.pad.a
8: -us sagus
9: _dEn-lil-ka e.Enlil.(a)k.a(k)
10: illgiI-zid digir.zid
1 1 : dUtu-kalam-ma-na Utu.kalam.anLa(k)
12: lugal-kalag-gi! lugal.kalaga
13: ...!
I !gfL- ' 5ki-ma
I Unm lugal.Urim.a(k)
14: lugal-an-ub-da-limmuzba-ke4 lugal.anub.da.(k) limmu.bLak.e
15: Kar-zid-da-i! Karzida.a
16: ud-ul-li-i!-ta ud.uli.a.ta
17: -par-bi nu-du-am gipar.bi.0 nu.(i.)du.0.a.am
18: en nu-un-til-Ia-am en.0 nu.(i .)n.ti1.0.a.am
1 96 Lesson 1 7

19: dAmar- dZuen Amarzuen

20: ki-g-dNanna-ke4 kLaga.a.Nanna.k.e
2 1 : -Qill:-kug-g-ni gipar.kug.ani.0
22: mu-na-du mu.na.(n.)du.0
23: en-ki-ag-ga-ni
- - - - -
en.kLaga.a.anL0
24: mu-na-ni-kur9 mu.na.nL(n.)kur.0
25: dAmar-dZuen-ke4 Amarzuen.(a)k.e
26: ud im-da-ab-sud4-re6 ud.0 1.b.da.b.sud.e.0
27: nam-tll-Ia-ni-e nam. til.anUe
28: -mu-na-ru a.mu.na. (n.)ru
Translation
1: For Nanna
2: of Karzida,
3: his beloved king -
4: Amar-Sin,
5-7: proposed by Enlil in Nippur,
8-9: patron of the temple ofEnlil,
10: the effective god,
1 1: the sun-god of his land,
1 2: the mighty king,
13: the king of Ur,
1 4: the king of the four quarters -
15: in Karzida -
1 6: from of old -
17: - its giparu not yet having been built -
1 8: - and no en having taken up residence in it -
19: Amar-Sin,
20: the beloved of Nanna -
22: built
21: his pure giparu.
23-24: He made his beloved en-priestess enter it.
25-26: Amar-Sin will prolong (its) days.
27: For the sake of his long life
 Lesson 1 7 197

28: he made a votive offering.

Commentary
1-2. Presumably this is a genitive phrase. A parallel construction occurred in Text 12:
dBiI-g-me En-dim-gigki illgal--ni, "for Gilgamesh of Endimgig, his king".
1 0- 1 1 . These two epithets did not occur in the previous inscriptions of Amar-Sin. Amar
Sin is not just content with the use of the divine determinative before his name; he must
also refer to himself as "effective god" and as "sun-god of his land". (Shulgi is also
referred to as digIT kalam-ma-na.)
14. The .e at the end is the ergative case-marker.
15. The writing in - is ambiguous. The original editor of this inscription interpreted it as
an anticipatory genitive: "of Karzida, its giparu" ) "the giparu of Karzida". However, it
may also represent a locative case-marker: "in Karzida". The .n in line 18 below would
seem to favor this latter interpretation.
16. Literally, "from a distant day", that is, something like "from of old", "since time
primeval". ta is the marker of the "ablative" case. This case can usually be translated as
"from", e.g., illgal-ta, "from the king". It is very common in the formation of adverbial
phrases.
The ablative is usually cross-referenced in the verbal prefix chain by the ablative
dimensional-prefix ta. However, in formulaic adverbial constructions such as this one, it is
not normally cross-referenced.
This particular expression is not uncommon. Outside of Ur Ill, for example, it occurs
in a Gudea inscription: ud-ul-li-i!.-ta numun--i!.-ta, "from of old, from when seed (first)
came forth" , "to go out", Akkadian wasill. ud-ul is also used in the formation of other
adverbial phrases, e.g., ud-ul-Ia-e ud.u1.a.M, "for a long time", "forever".
=

17. All the verb forms seen up to this point have been in what we would call in English
the "active voice". There has been a long discussion about whether or not Sumerian has a
"passive voice". Some Sumerologists say that Sumerian has no passive voice. Others say
that Sumerian is basically passival in nature.
To some degree this is a question of linguistic theory, and not of Sumerian. It is a
question of the definition and nature of active and passive, of the contrast between passive
and intransitive, and of the way such distinctions are marked in the morphology and in the
syntax. The problem is exacerbated by a tendency in the past to transfer categories found in
Indo-European or Semitic to Sumerian.
Some linguists would say, for instance, that the contrast active - passive does not exist
in ergative languages. Karl Oberhuber, for example, in his examination of the Sumerian
passive, has said that: "Ein eigentliches 'Passivum' ist dem Sumerischen als einer
Ergativesprache von Haus aus Fremd" (1982: 133). Earlier, Diakonoff said that ergative
languages "have no grammatical direct object, from which follows that ( 1 ) no Accusative
can exist; (2) no Passive and Active voice can exist" ( 1965 : 1 8). However, Sumerian is
only split ergative; this means that such theoretical constraints may not apply equally to the
maru and to the gamt!! (although not everyone agrees about the degree to which Sumerian
198 Lesson 1 7

is split-ergative).
The question also hinges on the nature of the Sumerian verbal root. Most
Sumerologists now say that the Sumerian root is unmarked for voice or transitivity; that is,
du can be active or passive, transitive or intransitive. These categories are not marked at all
in the root; rather, they are determined by the syntax and semantics of the entire sentence
within which they occur. The parallel has been made with English sentences of the type
"he is cooking". This sentence, on the surface, can either mean "the man is cooking-up",
because of the temperature, or "the man is cooking some food".
In any case, it seems that the two constructions which are differentiated in English as
"intransitive" and "passive" are expressed by one construction in Sumerian. (This is one
reason why some Sumerologists say that Sumerian has no passive.) Therefore, some
Sumerologists use the compound term "intransitive-passive" (or "passive-intransitive") to
refer to both constructions. Other Sumerologists use the terms "passive" or "intransitive",
based on how the corresponding construction in English (or German) comes out.
The difference between transitive and intransitive verbs can be illustrated using du, "to
build", and gin, "to go". In the bamty, these are:
(1) The king went.
luga1.0 i.gin.0
(2) The king built the house.
lugal.e e.0 mu.n.du.0
In the intransitive sentence, the subject of the intransitive verb (the patient) is marked by .0.
This is cross-referenced by the .0 at the end of the verb. To express "the house was built",
which in English would be called a passive, Sumerian uses a construction identical with
sentence ( 1 ):
(3) The house was built.
e.0 Ldu.0
The .0 case-marker of the subject of the passive verb is cross-referenced by the .0 at the
end of the verb.
Sentence (3) is essentially the construction seen in line 17, although with the negative:
(4) gipar.bi.0 nu.(L)du.0
nu is the same negative marker seen in Text 1 6. There, it occurred before a nominal
form (an active participle). Here it occurs before a verbal form. As stated in Lesson 1 6, nu
fits into the category of modal-prefix. As such, it is regularly followed by one of the
conjugation-prefixes. Here the conjugation-prefix (.i ) has assimilated into the lul of nu.
The enclitic copula has occurred several times, e.g., in the PN dBa- -nin-am, "Baba
is queen". In addition to being used in such equational sentences, am can also be used to
express circumstantial clauses. These can usually be translated into English as: "it being
that", "it being the case that", or by a participial phrase in -ing. When am is used in such a
construction, it must follow a nominalized sentence. Hence the verb form is to be
understood as: [nu.(L)du.0].a.am, meaning "it being the case that its giparu had not (yet)
been built".
To sum up, am is the enclitic copula, used here to express a circumstantial clause. .a
nominalizes the preceding sentence. The underlying sentence which has been nominalized
 Lesson 1 7 199

is: gipar.bi.0 nu.(i).du.0.

1 8 . The syntax of lines 17 and 1 8 is the same: am, following a nominalized sentence,
fonning a circumstantial clause. In line 1 8, tH (intransitive in English) is construed just like
the verb du (passive in English) in line 1 7:
[gipar.bi.0 nu.(L)du.0].a.am
[en.0 nu.(i).n(i).ti1.0].a.am
The writing is slightly different: ... -du-am vs. ... -tll-la-am. This is presumably be
cause du ends in a vowel, but tU in a consonant.
The verb fonn in line 1 8 differs in the presence of .n immediately before the verbal
root. This .n is somewhat difficult to explain. It cannot be the personal-affix .n, because tU
"to live" is intransitive here: "no en-priestess had yet taken up residence" in the giparu.
More likely, this .n is a reduced fonn of the dimensional-prefix which cross-references
the locative case. The usual form of this dimensional-prefix is ni. However, sporadic
instances of 1nl instead of Inil are attested; a possible case also occurs in Text 22. It has
not yet been possible to determine any phonological or morphological rules governing the
distribution of Inil and 1nl. It has been speculated that it has something to do with word
stress.
Just as the comitative dimensional-prefix da often occurs in a verbal prefix chain
without any corresponding comitative noun phrase for it to cross-reference, the locative
dimensional-prefix ni can also occur in verbal forms, without any corresponding locative
noun phrase. According to the studies of Gragg, such a use of ni often conveys a vaguely
adverbial force, and it can be translated simply as "there". It can refer loosely back to some
noun mentioned earlier, even if that noun is not in a locative case. In line 18, the reference
is to the giparu, the residence of the en-priestess, even though gipar.bi is not in a locative
phrase (and, in fact, is more closely connected syntactically with the verb fonn of line 17
than with the verb fonn of line 1 8).
20. The .e is the ergative case-marker. In line 14, an ergative case-marker .e already
appeared. Since in general in Sumerian, case-markers appear at the end of a nominal
phrase, no matter how long the phrase may be, the presence of the ergative case-marker in
line 20 makes the presence of the ergative case-marker in line 14 "un grammatical".
However, it is very easy to see how this situation arose: lines 4-14 fonn a logical unit,
consisting of nothing but epithets of Amar-Sin. Lines 15-18 are circumstantial clauses,
setting the background for Amar-Sin' s activities. Lines 19-20 are almost a parenthetical
addition, a sudden shift of topic back to the agent of the sentence, repeating Amar-Sin's
name. Once these lines were "inserted", it was only natural (and perhaps necessary?) to
again add the ergative case-marker .e. In fact, some copies of this inscription also have an
ergative case-marker .e at the end of line 9. The scribe may have unconsciously been
influenced by the fact that lines 1-14 fonn a complete unit by themselves. In several royal
inscriptions, the phrase lugal-an-ub-da-limmu2-ba-ke4 marks the end of an agentive nomi
nal phrase. In fact, if lines 15-20 were omitted, the remaining lines 1 - 14, 21 -22 would have
fonned a complete text by themselves.
2 1 . This is the direct object (patient) of the verb in line 22, "his (Nanna's) pure giparu".
22. This is the main verb governing lines 1-2 1 . The essence of these first 22 lines is:
200 Lesson 1 7

(1) For Nanna benefactive

(4) Amar-Sin agent
(15) -nothing having been built- circumstance
(19) Amar-Sin agent
(21 ) his giparu patient
(22) built. verb
24. As discussed above, the verbal root in Sumerian is not marked for transitivity; that is,
it can be used either transitively or intransitively, with no morphological change of the root
itself. For example, gyb (Text 6) can mean "to stand", or "to plant (a garden)". Thus, kur9
can either mean "to enter", or, as in this particular sentence, "to cause someone to enter", "to
bring in". The sense here is of installing the en-priestess in the giparu.
na is the dimensional-prefix which cross-references the dative, although no dative
phrase occurs. It is probably a kind of ethical dative or benefactive, loosely referring back
to Nanna, who is mentioned in lines 1 and 20.
ni is the full form of the dimensional-prefix cross-referencing the locative. Just as in
line 18, there is no locative noun phrase for it to cross-reference. It can be translated as
"there", referring loosely back to "his pure giparu" of line 2l.
In a sequence of two (or more) dimensional-prefixes, the dative dimensional-prefix
always comes first.
25. It is writings such as these which show that a personal name of the type "Amar-Sin" is
indeed a genitive phrase. The writing presumably stands for amar.Zuen.ak.e. As discussed
in Lesson 1 , there is evidence that in certain (undetermined) conditions the .ak of personal
names was lost, and so the name was transcribed as "Amarzuen". Jacobsen, on the other
hand, even in his historical and literary publications, refers to this ruler as "Amarsuenak".
26. The four most common conjugation-prefixes in Sumerian are mu, L ba, and bi ; ex
amples of all of them have occurred. Besides these four, there are a certain number of
others, all with a Im/. The two most common are: im-ma and im-mi, with reduplicated
Im/. Others are written with one Im/: l-mi and l-ma. Others occur with different initial or
final vowels: am-ma.
The relationship among these forms is unclear. It is not sure if these variations are
(mostly) orthographic (im-mi - i-mi), mostly phonological (im-ma - am-ma), or correspond
to a difference in meaning. These forms will be discussed further below.
In addition to these bi-syllabic forms, a form written im also occurs; this is what is
written in line 26. Some Sumerologists believe that im is a reduced form of im-ma or im
mi (although the conditions governing such reduction are unknown). However, most
Sumerologists think that a form such as im-da derives phonologically from libdal Lb.da;
=

this is the analysis followed here. In Text 17, for example, there occurs: im-da-ab-sud4 -
re6. Under this assumption, this would derive from: Lb.da.b.sud.e.0. .i is the con
jugation-prefix, and .b is the optional pronominal-prefix which appears before the
comitative dimensional-prefix; here the reference is loosely ud, which is inanimate. The
phonological change may have been along the lines of I ibdal ) I iddal ) I imda/.
This explanation of im is plausible, because it is indeed normally followed by da or ta.
 Lesson 1 7 201

If correct, then it should be understood as basically the conjugation-prefix i. There are

cases, however, where im is not followed by da or ta, and where some other explanation
may be necessary.
da is the comitative dimensional-prefix. It often occurs with no corresponding comi
tative nominal phrase. Here it loosely refers back to ud, even though the latter is actually
cross-referenced as a patient.
Since the verb form is man1, the .b before the verbal root cross-references the direct
object ud. The .0 after the man1-suffix cross-references the subject Amarsin.ak.e:
Amarsin.ak.e ud.0 Lb.da.b.sud.e.0
As was just said above, the ud is directly cross-referenced by the .b before the verbal
root, but it is also loosely cross-referenced by the comitative dimensional-prefix da, with its
(assumed) inanimate pronominal-prefix .b.
sud4 is a member of the affixation class, forming its man1 by addition of the man1-suf
fix .e.
To summarize the verb form:
i . b . da b . sud . e . 0
(1) (2) (3) (4) (5) (6) (7)
(1) conjugation-prefix
(2) optional pronominal-prefix
(3) comitative dimensional-prefix
(4) personal-affix cross-referencing patient
(5) verbal root
(6) maru-suffix
(7) personal-affix cross-referencing agent
Discussion: structure
The bare-bones structure of this text is:
1 -3 For Nanna benefactive
4- 14 Amar-Sin agent
15 in Karzida place
16 from of old time
17 no giparu having been built circumstance
18 no en having lived circumstance
19-20 Amar-Sin agent
21 his giparu patient
22 built. verb
23 His en patient
24 he made enter. verb
25 Amar-Sin agent
26 its days patient
26 will prolong. verb
202 Lesson 1 7

27 For his life purpose

28 he dedicated. verb
- Passive
Ergativity is defined by the way the relationships among the primary participants in a
sentence are marked. In the bamt!!, which functions on an ergative basis, the subject of a
transitive verb is marked by .e. The subject of an intransitive or passive verb, and the direct
object of an active verb, are marked by .0. (The man} of intransitive-passive verbs will be
discussed later.)
In Sumerian,there is no difference in construction between an intransitive sentence and
a passive sentence in the bamt!!:
luga1.0 Lgin.0
e.0 Ldu.0
Perhaps the most cautious statement about the passive in Sumerian is that of Jacobsen:
The external criteria detennining whether a Sumerian fonn is active or
passive in meaning are as yet far from clear and the whole question whether
in actual fact this distinction may be considered gennane to the Sumerian
verb is yet to be decided (1956:49*).
Similarly, Viktor Christian, many of whose ideas about S umerian grammar are outside
the main-stream, says:
In Sumerian, we do not find the categories "transitive" and "intransitive", or
"active" and "passive". The fact that we are often forced to translate as active
or passive, transitive or intransitive, only results from the inadequacies of our
own language to correctly reproduce S umerian (196 1 : 1 3).
That is, the contrasts active - passive, and transitive - intransitive, may not be the most
fitting way to describe Sumerian. Some Sumerologists have stated that an analysis in tenns
of action - state would more fittingly describe Sumerian. In practice, however, people
sometimes tend to be less than rigid in their use of such linguistic tenns.
Traditionally, Sumerian grammars present two paradigms: one for the active; and one
for the intransitive-passive. There is no difference in the structure of the root; rather, the
differences lie in the function of the personal-affixes, and in the way that the case-endings
of the primary participants are cross-referenced. It may be useful to summarize the bamm
fonns in the singular.
active intransitive-passive
first person mu.0.sar mu.sar.en
second mu.e.sar mu.sar.en
third animate mu.n.sar mu.sar.0
inanimate mu.b.sar mu.sar.0
- Conjugation-prefixes
The relationship of the conjugation-prefixes im-mi, im-ma, im, and other rarer fonns is
unsure. It is usually assumed that these are not unanalyzable, unitary morphemes. Rather,
 Lesson 1 7 203

many Sumerologists assume that the two conjugation-prefixes im-mi and im-ma somehow
derive from bi and ba (although it is not always made clear if this derivation is to be
understood in synchronic or diachronic terms). Falkenstein, for example, who has a
different understanding of the conjugation-prefixes than that presented in this book, derives
im-mi from *i-bi. Others, however, have argued for the existence of a morpheme fmf in all
these forms. It would roughly correspond in function to the Akkadian ventive (which is
normally used with verbs of motion, marking movement towards the speaker).
Such historical developments are unproven, and perhaps unprovable; Falkenstein, for
example, had to posit a number of apparently unmotivated phonetic changes to get his
forms to work. In general, some Sumerologists (such as Falkenstein) are inclined to see
variation in Sumerian as due to phonetic reasons, even if the rules governing the phonetic
changes cannot be determined; others (such as Jacobsen) are inclined to believe that
Sumerian grammar is more complex than usually believed, and that the variation we see is
due to our ignorance of the morphology, not to unexplained phonetic accidents. In any
case, in synchronic terms im-ma and im-mi (and im) pattern the same as the other conju
gation-prefixes; that is, they all occupy the same position in the verbal prefix chain, and
their presence is mutually exclusive. More work remains to be done on the synchronic
distribution of these conjugation-prefixes. Krecher (1985) is the latest attempt to isolate the
morphology and the semantics of the various conjugation-prefixes in fmf. He posits a
rather wide range of functions, which need to be more fully investigated.
- Case-markers
The presence of the two ergative case-markers (lines 14 and 20) has parallels in other
agglutinative languages. If a construction starts to become very long or convoluted, the
speaker (or writer) will occasionally "get lost" in the construction, and may occasionally
back-track, changing the topic, and will have to repeat a previous case-marker.
- History
The giparu at Ur was the "official" dwelling-place of the en-priestess (who is some
times referred to by the Akkadianized term "en tu-priestess"). It was a large structre,
composed of many rooms. The first such structure at Ur may go back to Early Dynastic
times; it was built and rebuilt right through the Neo-Babylonian period.
The en-priestess was always of royal blood. Perhaps the most famous was
Enheduanna, the daughter of Sargon of Akkad. She is well-known as the author of two
well-preserved poems, written in good Sumerian. In the Neo-Babylonian period, Nabo
nidus installed his daughter in the position.
The en-priestess "represented" the goddess Ningal, in some way. In particular, she
represented the goddess Ningal, while the reigning monarch represented the god Nanna
(the husband of Ningal) in some kind of "divine marriage ceremony". This ritual may have
originated as an end-of-harvest-time festival. It has often been discussed among Sumero
logists, and there has been much disagreement about what the sacred marriage was meant to
represent.
Penelope Weadock has summarized the functions of the giparu:
204 Lesson 1 7

Three separate units emerge from the Ur III--Isin-Larsa giparu building: the
Ningal temple which is the locale in which the entu-priestess, as the incar
nation of the goddess Ningal, carried out her most important function as a
participant in the rite of the sacred marriage; the giparu proper which was the
official dwelling of the entu-priestess, with its annexe, the cemetery for the
former entus; and the sanctuary in which the entu prayed for the life of the
king, her father or brother, in the hope that the gods would bestow prosperity
upon the land through the king, their human regent. (1975: 124).
It has been argued that the "institution" of the giparu existed in other cities, for ex
ample, in Uruk. This is probably true, but the evidence is not conclusive. However, Tex1
17 evidently refers to a giparu in Karzida, not to the giparu in Ur. This implies a Nanna
temple in Gaesh. Nothing is known of this temple or this giparu.
The original meaning of the term -par is unknown; it may have been "storehouse'
of some kind.
 Lesson 1 8
This is a door socket of Shu-Sin, the brother and successor of Amar-Sin. He ruled
from 2037 to 2029 BC.
Sign-list and vocabulary

An-nu-ni-tum Annunitum (DN, masc)

* 17u>
i *-Jj Su-dZuen Shu-Sin (PN)

dam wife, consort

Notes

An-nu-ni-tum This may originally have been an epithet of Inanna. In Sargonic texts, the
d
divine name lnanna-An-nu-ni-tum is occasionally attested. But after the Old Akkadian
period, this compound term does not occur, only the individual term An-nu-ni-tum. Ac
cording to Jack Roberts, this pattern of attestation "suggests that the epithet split off and
became an independent deity" ( 1972: 147).
Gelb has pointed out "the tremendous number of compound divine names in the Ur III
period" ( 1987: 125). The name dlnanna-An-nu-ni-tum belongs to the class of names cha
racterized by Gelb as "DN plus description".
The etymology is unknown. It is possible that the -itum ending is an Akkadian femi
nine gentilic: II-t-um/. However, the meaning of "anun" or "annun" is unknown; it is not
sure ifit is Akkadian or Sumerian.
S u-dZuen The name is Akkadian, meaning "the one of Sin", or "the one belonging to
Sin". It was formerly read as "Gimil-Sin".

205
206 Lesson 18

Text 1 8

1. Th e name of the goddess is clearly A n -nu-ni-tum.

 Lesson 18 207

Notes

Many objects containing inscriptions have been damaged, either in ancient times or in
modern times. This means that part of the inscription may be completely broken away, as
in the bottom right-hand corner of Text 17, or part of it may be effaced, as is the top left
hand corner, and as in case 8.
It is important that transliterations reflect such damage, especially when the text is
completely broken. Unfortunately, there is no one system in use by Sumerologists (or
Assyriologists) to indicate such breaks. The most common system is to use brackets.
These are used to include signs completely broken away. Thus, line 1 1 is best trans
literated: --[ni]. Brackets can also be used to indicate partially broken signs. For
example, line 8 can be transliterated: lugal-[kala]g-g, and the last line as: mu-n[-du] .
Brackets are imprecise, however, in such cases as the partially effaced An-sign in line
. To resolve this problem, some scholars (not all) use half-brackets to indicate partially
broken signs. Thus, the first sign of line 1 can be transliterated rAn' .
Instead of half-brackets, some scholars use dots under vowels, to indicate that the sign
is partially broken or effaced. Thus, this An-sign might be transliterated as An. Some
scholars use different combinations of full brackets, half-brackets, and dots, in sometimes
rather idiosyncratic combinations.
The decision about whether or not to use brackets (or half-brackets) is not always
clear-cut. For example, what about the turn-sign in line 1 , or the first part of the Urims-sign
in line 9? In practice, such damage to the text is often ignored, if the context and the
remaining traces of the sign make the sign unambiguous.
Similarly, it is difficult to decide how breaks should be reflected in translation. Some
Sumerologists use brackets in translation, reflecting the breaks in the text. However, since
English and Sumerian are of such different grammatical structures (particularly in word
order), this procedure can be cumbersome and tiresome. Many scholars, therefore, omit
brackets i translation, especially when writing for a professional audience, because such
an audience will be able to follow or control the transliteration or autograph.
Occasionally, scholars may be suspicious of a published transliteration, or even an
autograph, of a cuneiform text. Or, they may wish further information about a partially
effaced sign. In such cases, they may collate the text, that is, physically examine the
cuneiform document. If the text is not easily accessible, they may ask another scholar to
perform such a collation.
The system used here is the simplest. Full brackets are used only to indicate
significant breaks. They are omitted from transcription and from translation.
Transliteration Transcription Translation

1 : An-nu-ni-tum Annunitum For Annunitum,

2: dam--ni-ir dam.anLr his wife -
3: dSu-dZuen Susin Shu-Sin,
4: ki-2-dEn-lil-1{1 kLaga.a.Enlil.a(k) the beloved of Enlil,
208 Lesson 18

5: lugal-dEn-lil-le lugal Enlil.e the king whom Enlil

selectea in his
6: ki-2-Sag4-g-na kLaga.0.Sag.anLa loving heart,
7: in-pad Ln.pad
8: luga1-[kala]g-g lugal.kalaga the mighty king,
9: lugill-Urimski-ma lugal.Urim.a(k) the king of Ur,
10: lugal-an-ub-da-limmuZ lugal.anub.da.(k) limmu.bLak.e the king of the four quarters-
oa:}(e4
1 1 : --[niJ e.anL0 her temple
12: mU-!l@-.ill!I mu.na.(n.)du.0 he built.
Commentary
1 . Lines 5-7 are an epithet of Shu-Sin, which occurs in virtually all of his inscriptions. It
was not used by his predecessors, nor by any subsequent ruler.
Although this epithet is very common, and its basic meaning is fairly transparent, the
syntax underlying it is not clear.
In particular, the grammatical relationship ofki-g to Sag4 -g-na is not sure. One in
terpretation of these lines is: "the king whom Enlil has elected as the beloved of his heart",
that is, kLaga.a Sag.anLa(k). However, Sumerian would probably use -Se to express the
complement introduced in English by "as"; cf. Text 19, where -Se is used in a (roughly)
parallel context.
Another interpretation is to see ki-2 as an active participle modifying Sag4 ' with the
entire noun phrase being in the locative: "in his loving heart", that is, [kLaga.0 Sag.ani].a.
The problem with this interpretation is that modifiers of nouns almost always follow their
nouns, not precede them; one would not expect kLaga.0 to precede Sag.anL However,
there are exceptions to this rule, and there are even occurrences where simple adjectives
precede their nouns (although these occurrences are mostly of a formulaic nature). There is
probably some stylistic emphasis present.
Lines 5-7 consist of the noun lugal and a relative clause. Since this is a relative clause,
one would expect to find a sentence nominalized in .a, as was the case in all other relative
clauses with a verb. However, this common epithet is always written in-pad, with no
nominalizing .a. This is difficult to explain. It is not simply a problem of orthography. A
follower of the Falkenstein school might wonder about the possibility of a reading /pada/
for the pad-sign. However, no such value is recognized by the standard sign-lists, and in
any case, one would expect it to be written in-pad-da (as in Amar-Sin's epithet, Enlil.e
Nibru.a mu.pad.a, always written mu-pad-da).
Somewhat similar instances of relative clauses without an expected nominalizer occur
elsewhere in S\lmerian, sometimes of a formulaic nature. They need further investigation.
7. i is the conjugation-prefix. n. is the personal-affix cross-referencing the U.run.t!!-agent,
Enlil.e.
 Lesson 18 209

Discussion: structure
The structure of this text is:
[Annunitum, dam.ani].r benefactive
[Susin, ki.aga.a.Enlil.a(k), agent
lugal Enlil.e ki.aga.0.ag.anLa tn.pad,
lugal.kalaga, lugal. Urim.a(k),
lugal.anub.da.(k) limmu.bLak].e
[e.ani].0 patient
mu.n.(n.)du.0 verb
It thus follows the pattern of most royal inscriptions. This basic pattern, however, is
somewhat difficult to recognize, because of the length and the complexity of the
appositional phrases in lines 4 through 1 0.
-Relative clauses
The interpretation of lines 5-7 given above follows Jacobsen, who translates this
formulaic phrase as "the king whom Enlil envisaged in his loving heart". He considers it to
be a clause nominalized "in zero"; however, he cannot find many close parallels. These
lines illustrate a problem encountered several times before: a construction occurring fre
quently, its meaning relatively transparent, but its syntax dubious. There are still several
problems in the understanding of Sumerian relative clauses.
- Adjectives
There are a few cases in Sumerian where adjectives (or other modifiers) precede their
head noun, instead of following it. For example, the adjective kug, meaning "pure", quite
frequently precedes the names of gods and goddesses: kug-dInanna, "pure Inanna". It ap
pears to be the only adjective to be used so regularly in this position.
Other languages whose order is basically noun-adjective, such as French, also permit a
certain number of cases of adjective-noun constructions. These cases are usually limited to
a fixed number of adjectives or expressions. In general, it seems that languages of the
noun-adjective type permit more exceptions than do languages ofthe adjective-noun type.
- History
The circumstances of Amar-Sin's death are unclear. A late omen text says that he died
of an infection caused by a foot bite of some kind. Shu-Sin was his son and successor
(although it has been said that he was his brother, not his son). It was during the latter's
reign that trouble began to be felt in the empire: the Sumerians began to feel the pressure of
the Amorites.
Curiously, a number of "love poems" (to use Jacobsen' s term) have been preserved,
most of which are directed to the fourth king of the dynasty, Shu-Suen. One
guesses that this king, or perhaps more likely his queen, had in his entourage
a woman poet who enjoyed singing about love and lovemaking, and whose
210 Lesson 18

works, since they were cast in the fonn of praise for the king's beauty and
virile prowess, were favorably received and carefully preserved in writing
(1987:85).

Text 1 8 a
supplementary

This is a votive bead of carnelian, found at Susa. It has been speculated that this bead
was carried off to Susa as booty by the Elamites when they sacked Ur in 2004 BC, putting
an end to the Ur III Dynasty.

* llfl 1 1 \\
W $_ i fl1 l > l=I \-
 Lesson 1 9
This door socket of Shu-Sin records one of the more significant events of his reign.
Sign-list and vocabulary

Sara2 Shara (DN, masc)

Nin-lil
- -
Ninlil (DN, fern)

the Amorites; the West

}---<{ JP Mu-n-iq Ti-id-ni-im Muriq-Tidnim (GN)

j <!i> <'1 s=:r e-ag4 -g-ru\Q-da Eshagepada (TN)

jIj nir-gaI prince

5=J ad-da father

1-- gib (kind of priest)

}-- gudug (gudu4) (kind of priest)

u hand

q::ffi sipad (sipa) shepherd

ne forces, troops

8=J ma-da land

o <; dadag to be clean, pure

gal great

Notes

Sar God of the city of Umma; son of Inanna. Not much is known of him; he did not
rank very high in the Sumerian pantheon.
Nin-l1l According to Sumerian mythology, Ninlil was raped by Enlil, and later became his
wife (darn). Nanna was their child. She was especially worshipped in Nippur.

21 1
212 Lesson 19

Mar-tu This term was used by the Sumerians in two ways. In an "ethnic" sense, it refers
to the (West-Semitic) Amorites, who dwelt in the Syrian desert west of Mesopotamia. It is
also used in a rather vague geographical sense, meaning "west" in general.
The etymology is unknown. Some now read it as Mar-du.
Mu-ri-iq Ti-id-ni-im The name is Akkadian, Muriq Tidnim, meaning "that which keeps
Tidnum away". muriq is the D-stem active participle in the construct state, from req!!, "to
be far away". Tidnum (or "Didnum") is probably the name of a particular Amorite tribe,
although here it refers to the Amorites in general. In late lexical texts, "Tidnum" is equated
with the Akkadian word used for the Amorites: ri-id-nu !!-mur-ru-y. The original form
=

and etymology of the name are unsure. In the various Semitic languages, several variant
spellings of the name Tidnum occur: Ti-da-nu-um and Di-ta-nu-um in Akkadian, Ddn and
Ttn in Ugaritic, etc. It is not always sure if such terms refer to the same people.
E-ag4-ge-pAd-da The principal temple of Shara in Umma. The name means "the temple
chosen in (his) heart", that is, e.ag.e.pad.a. The .e is the mark of the locative-terminative
case.
nir-g8.l It is difficult to determine the precise meaning of such a word; it is conventionally
translated into English as "prince" or "hero". The Akkadian equivalent is etellu, glossed by
the CAD as: "prince, lord".
In origin, this is probably an active participle with an incorporated object. gal normally
means "to be", but can also mean "to have". The combination nir-gal would mean "the one
who has nir". Unfortunately, it is not known what nir means.
ad-da Sumerian has three words for father: !!-!!,ab-ba, and ad-da. Lambert (1957) stu
died the distribution of the three terms on (primarily) geographical and class lines, but could
not come to any definite conclusions about their usage.
iib This is a very old loan-word into Sumerian from Akkadian waSipu. This Akkadian
term is usually translated "exorcist". iSib was then loaned back into Akkadian as iippu,
glossed by the CAD as "purification priest". iib itself is often translated as "incantation
priest".
It is never easy to determine the exact function of any priestly office. Iohannes Renger
(1967f) has studied the Old Babylonian priesthood, and exhaustively discussed the
Akkadian words for the different kinds of priests.
gudug The equivalent Akkadian priest is the paSgu, from the root paSau, meaning "to
anoint". The gudug-priest seems to be especially involved in certain kinds of ritual
activities, which it is not yet possible to define exactly. He was of a lower rank then the
iib-priest (at last in the Old Babylonian period); temples could have several gudug-priests,
but apparently no more than one iSib-priest.
This word is often transliterated as guda4 ' with the second vowel as lal, not as lu/.
sipad The cuneiform sign used to represent this word is actually a combination of two
 Lesson 19 21 3

signs: the Qf!-sign ( * ) and the udu-sign <,ffi) . The Qf!-sign was originally a picture of a
staff or sceptre of some kind; in this meaning, it is read as gidri. It was also used to
represent the word for "overseer", Le., a man holding a staff of authority; in this meaning, it
is read as !!gula. The udu-sign means "sheep". Thus, the sipad-sign graphically represents
approximately "the overseer in charge of sheep".
It is reasonably sure that this word has a Id/-Auslaut. However, it is much more
common to find it transliterated as sipa.
ma-da This is usually assumed to be a very early loan into Sumerian from Akkadian matu
(although the latter is itself of dubious etymology; the only other Semitic language it occurs
in is Aramaic). The three terms kalam, ma-da, and kur are often translated into English as
"land", but they are not synonymous. Limet (1978) has studied their distribution, espe
cially in documents from the Ur III period. kalam is used exclusively to refer to Sumer.
kur originally meant "mountain". It then came to mean "foreign land"; it is never used to
refer to the land of Sumer. ma-da is more problematical; the sense is approximately
"territory". It is used mostly for foreign lands, but in certain uses it can refer to Sumer.
Jacobsen believes that in contexts such as Text 19 ma-da means "steppe". The opposition
between kalam and kur has also been studied by Steiner (1978) on a number of levels:
historical, legal, etc.
Because the term kalam is restricted to the land of Sumer, Poebel has speculated that
the word Ki-en-gi was a dialectal form of kalam. ka-na-.M is used as the Emesal equivalent
of both Ki-en-gi and kalam. However, in earlier texts kalam may have had a more general
meaning. Raphael Kutscher says that "Although the literal meaning of kalam is 'country'
(matum), it narrowed its scope to 'The Country' par excellence, namely, Sumer, and even
tually, to 'the nation', Le., the Sumerians" (1975:68).
dadag The reading is not certain; it is variously read as: da7 -dag, dag-dag, zalag-zalag,
zalzalag, and babbar2 ' The reading as I dadagl seems established by late lexical texts,
which give the syllabic writing da-da-f!g as the equivalent of the Akkadian word for "pure",
ebbu.
The dadag-sign is formed by the writing of two <) signs; the situation is similar to
that of the siskur1 -sign. The word Idadagl is probably a reduction of something like
*/dagdag/, a reduplicated form. The reduplication may have had some kind of intensive
value in its origin. This explains why there are two signs. Originally, one sign stood for
Idag/; two signs stood for Idagdag/. Later, */dagdagl was reduced to Idadag/, but the
original two signs continued to be written. Similarly, it has been speculated that the word
Isiskurl was originally a reduplicated form, perhaps */sikur-sikur/; this explains why it
was written with two signs. One would then have to assume a development along the lines
of */sikursikur/ ) */sisikur/ ) Isiskur/, or something similar.
2 14 Lesson 19

Text 1 9

15

BRIt>
o t> R est m istake ofscribe.
 Lesson 19 21 5

Notes

Line 23: As the editors point out, the scribe has drawn a ni-sign, instead of the expected
du-sign. He may have been unconsciously influenced by the ni-sign directly above, in line
22.
Transliteration Transcription

1: dSara2 Sara
-
2: nir-Ml-An-na
- - -
nirgal.An.a(k)
3: dumu-ki-g2 dumu.kLaga.a
4: dlnanna Inanna.(k)
5: ad-da-ni-ir adda.anLr
6: dSu- dZuen Susin
7: mb-An-na iib.An.a(k)
8: gudug-u-dadag gudug.u.dadag
9: dEn-HI Enlil
1 0: dNin-lil-ka Ninlil.(a)k.a(k)
1 1 : .!! dig.ir-gal-gal--ne u digir.gal.gal.ene.(k)
12: lugal dEn-lil-Ie lugal Enlil.e
1 3 : ki-g2 kLaga.0
14: ag4 --na ag.anLa
15: in-pad Ln.pad
1 6: sipad-kalam-ma-e sipad.kalam.a(k). e
17: lugal-kalag-g lugal.kalaga
Ski-ma
1 8: -I-ygI-Unm lugal.Urim.a(k)
19: lugal-an-ub-da-limmuZ-ba-ke4 lugal.anub.da.(k) limmu.bLak.e
20: ud bad-Mar-tu ud bad.Martu.(k)
2 1 : Mu-n-iq Muriq
22: Ti-id-ni-im Tidnim.0
23: mu-du- mu.(n.)du.0.a.a
24: .!! ne-Mar-tu u ne.Martu.(k.)0
25: ma-da-ne- mada.anLe
26: bi-in-g4- bLn.gL0.a.a
27: E-ag4-g-pad-da Eagepada
216 Lesson 19

28: -ki-g-M-ni e.kLaga.a.ani.0

29: nam-ti1-la-ni-e nam.til.anUe
30: mu-na-du mu.na.(n.)du.0
Translation

1: For Shara,
2: the prince of An,
3: the beloved son of Inanna,
5: his father -
6: Shu-Sin,
7: the iib-priest of An,
8: the gudug-priest with the pure hands
9: ofEnlil and Ninlil
1 1 : and of the great gods,
1 2: the king whom Enlil selected in his loving heart
1 6: to be shepherd of the land,
17: the mighty king,
18: the king of Ur,
19: the king of the four quarters -

23: when he built

20: the Martu-wall
21: (whose name is) Muriq-Tidnim,
26: and when he drove back
24: the forces of the Martu
25: to their own land -
30: he built -
27: theEshagepada,
28: his beloved temple -
29: for the sake of his life.
Commentary

8. dadag is an adjective from the root "to be pure".

The presence of the ka-sign in line 10 means that the construction is a double genitive:
"the gudug-priest of the pure hand(s) of Enlil and Ninlil, and of the great gods". It is
difficult to say exactly what this means. However, the "clean hands" of the paBu-priest
 Lesson 19 217

are explicitly mentioned several times in Akkadian texts. The sense of the sentence is
probably: "the gudug-priest with the pure hands, in the service of Enlil and Ninlil, and in
the service of the great gods".
9- 1 1 . Enlil and Ninlil are conjoined, without any conjunction. These two deities are set
off from the more vague "great gods" by the conjunction y.
The two plural formations of nouns seen so far are the morpheme .ene (for animate
nouns), and reduplication (for inanimate nouns). Another common plural formation, used
when a noun is followed by an adjective, is to reduplicate the adjective: digir-gal-gal. This
type of formation is most common with animate nouns, but a few examples occur with
inanimate nouns: inim-gal-gal, "the great words". In addition, it is possible to tack on the
plural marker .ene: digir.gal.gal.ene, which is the form in Text 19.
Plural formations such as digir-gal-gal and digir-gal-gal-:-ne are usually thought to be
superlatives of some type: "the most great gods". This is very hard to prove. It is
especially hard to determine if there is any difference between the reduplicated forms with
.ene and the reduplicated forms without it. Only a few adjectives seem to occur in such
plural formations; by far the most common is gal.
In Text 6, the plural "king of the gods" was written illgal-digir-re- with redup
lication of the final Irl of digir. Line 1 1 of Text 19 uses a more morphemic writing; this is
the regular writing of this particular expression.
1 6. One of the functions of the terminative case in -e is to express the purpose of certain
verbs. Here the sense is: "selected to be the shepherd of the land". Expressions of this
type are not infrequent in the royal inscriptions. In them, the nominal phrase marked in -e
regularly follows the verbal form in-pad, instead of preceding it.
This deviation from standard Sumerian syntax is presumably to give some degree of
emphasis to this last constituent. There are instances where adverbial phrases, and even
patients, are placed after the verb phrase, instead of before it.
19. The .e marks the end of the ergative agent phrase, spanning lines 6 through 19.
20. Lines 20-26 form two when-clauses, governed by the ud of line 20:
20: when he built ...
24: and when he drove back ...
30: (then) he built ... .
The two when-clauses are linked by the conjunction y in line 24. The syntax of these
clauses is the same as that seen in Text 12. ud is the relative marker; it is followed by two
clauses, each of which is nominalized in .a, and each of which is in the locative case. A
more literal translation would thus be: "on ( the locative .a) the day that ( the norninalizer
= =

.a) ... ":

ud [bad.Martu ... mu.(n.)du.0.a].a
u [ne.Martu ... bLn.gi.0.a].a
bad-Mar-tu presumably a genitive phrase, "the wall of Martu". Since Mar-tu can
is
also mean "the West", this is sometimes translated as "the Western Wall". What follows is
the actual name of the wall. Instead of using a construction with the word for "name" (mu),
the name is expressed through an appositive. Such formulations are quite common.
24. !l links the two temporal clauses, each dependent on the relative marker ud:
218 Lesson 19

ud: 1) ... mu-du-

2) ... bi-in-g4-f!
25 . . ani is somewhat ambiguous. It could refer back (loosely) to Shu-Sin, or it could refer
back to the ne-Mar-tu of line 24, treating the term as a singular or a collective.
.e is the marker of the locative-terrninative case. As discussed in Lesson 9, this case
shares some of the characteristics of the locative case marked in -, and some of the
characteristics of the terminative case marked in -e. This can lead to a certain amount of
ambiguity. For example, this particular line has been interpreted in two ways. One
interpretation is to understand .e here in the sense of -e. The meaning would then be: "he
drove the Amorites back to their own territory". However, it is also possible to understand
.e in the sense of -f!; the meaning then is: "in his own territory, he drove out the Amorites".
As was also the case in Text 9, the locative-terminative is not cross-referenced by any
dimensional-prefix in the verbal chain.
Line 25 has been transliterated here as ma-da-ne-. This assumes that the lit of lanil
has contracted into the lel of the locative-terminative case-marker, producing a pronun
ciation something like Imadane/. Other scholars have transliterated the line as ma-da-ni-.
Some Sumerologists do this, because they do not believe that such a contraction took place.
Others do it, because they are consciously being morphemic in their transliteration. This
problem is further discussed in Lesson 2 1 .
26. In the verb i n line 23, the personal-affix . n for the bamt!!-agent is not expressed in the
writing. However, in line 26 it is so expressed:
23: mu-du-f!
26: bi-in-g4-i!.
It is also not expressed in the verb in line 30:
30: mu-na-du.
From this one text, a facile generalization would be: The personal-affix .n is not used
when the conjugation-prefix mu is present, but it is used when the conjugation-prefix bi is
present. However, in other texts the writing mu-na-an-du occurred. It is such varying for
mulations which makes it very difficult to determine the rules governing the presence or ab
sence of the personal-affix.
Discussion: structure
Although this inscription is thirty lines long, it consists of only one sentence. The one
finite verb form is in line 30. It may help to summarize the structure of the text:
1 For Shara, benefactive
6 Shu-Sin, agent
20 - when he built ... - circumstance
30 he built verb
27 theEshagepada patient
29 for the sake of his life. purpose
 Lesson 19 219

It is not uncommon to find a verb with a bi conjugation-prefix co-occurring with a

nominal phrase in the locative-terminative case. In Text 19, line 26 uses the conjugation
prefix hi, and line 25 has a nominal phrase in the locative-terminative. Earlier it was
mentioned that it is also not uncommon to find a verb with a ba conjugation-prefix co
occurring with a nominal phrase in the locative case. The significance of these facts still
remains to be integrated into a coherent theory of Sumerian grammar.
- Loan words
Both i'ib and ma-da are early loan words from Akkadian. One, however, ends in lal,
and one doesn't. Gelb has speculated that such very early loans reflect a stage of Akkadian
when the case-system was not as fully developed as it was during the Classical periods.
However, it has also been suggested that the lal is the Sumerian nominalizer.
- Chronology
As the name of this wall implies, its function was to keep away the nomadic Amorites.
Individual Amorites, or small groups, had been entering into southern Mesopotamia from
the North-west for many years, but during the reign of Shu-Sin they began to enter in
force.
The building of this wall is also mentioned in a "year-date" of Shu-Sin. Until the
Seleucid period, there was no chronological system in Mesopotamia based on a fixed date.
Dating systems varied from place to place and from time to time. Beginning at least as
early as the Old Akkadian period, individual years in a king's rule were given their own
names. For example, the first year of the rule of Shu-Sin was named: "(the year when)
Shu-Sin became king"; his third year was named: "(the year when) Simanum was
destroyed". That is, the name given to the year referred to some important event in the rule
of the king. These year-names were gathered into lists. Without these ordered lists, it
would be impossible to know which particular chronological year a given year-name
referred to.
Year 4 of Shu-Sin's rule is named dSu-dZuen illgal Urims ki-ma-ke4 bad-Mar-tu Mu
ri-iq-Ti-id-ni-im mu-du, that is, "(the year when) Shu-Sin, king ofUr, built the wall against
the Amorites, (named) Muriq-Tidnim". The wording of this year-name is similar to the
wording in lines 20 through 23 of Text 19. Douglas Frayne has said: "It can be
demonstrated that temporal clauses in royal inscriptions of the Ur III through Old
Babylonian periods often allude to year formulae of the king" ( 1983:745); this is apparently
such an instance.
In a similar vein, Hallo has pointed out:
The correlation between neo-Sumerian regnal years on the one hand and
royal hymns on the other is a high one both in terms of numbers and in terms
of content. ... Is it too daring to suggest that each date formula was formally
introduced together with a new hymn? (1966: 1 39, and n.82).
The year-formula just cited is the "long" form of the year-name; there is also a "short"
form: bad-Mar-tu ba-du, "(the year when) the wall of Martu was built". This latter is a
220 Lesson 19

passive construction, bad.Martu.(k).0 ba.du.0.

Year-dates often occur in both a long fonn and a short fonn. The long fonn has an
agent marked in .e, and a verb with conjugation-prefix mu. The short fonn is without
agent, and has a verb with conjugation-prefix ba. For example, there are several year-dates
of the type: mu PN.e GN mu-bul, "the year when PN destroyed GN", and mu GN ba-bul,
"the year when GN was destroyed". In each case, the verbal sentence is in apposition with
the noun mu "year".
The conjugation-prefix ba frequently occurs with verb fonns in passive sentences,
such as the short fonn of year-dates. In the Old Babylonian Grammatical Texts, Sumerian
verbal fonns with the conjugation-prefix ba are usually translated by the Akkadian sterns
with infixed It I , that is, sterns with separative, reflexive, and passive meanings. However,
Malcolm Horsnell (1977) has questioned the standard interpretation of the short fonn of
year-dates as passives, arguing that they should be translated as agentless active sentences.
It has also been speculated that there is not just one conjugation-prefix ba, but rather
there are two; that is, they are homonyms. One is seen chiefly in passive sentences, the
other in less definable contexts. Needless to say, it is very hard to prove such an assertion.
-Literary parallels
Other references to this wall have been preserved. One is in a letter by Sharnim-bani,
the official in charge of its construction; he writes to Shu-Sin complaining of his troubles.
Shu-Sin's own reply to Sharrum-bani, in which he berates him for neglecting his duties, is
also preserved.
At first blush, it might strike one as rather astounding that such letters just happen to
be preserved. But the reason is because these (and other) letters came to be considered
literary texts, and were used for scribal practice. According to the definition of
Michalowski, "royal literary letters are thus simply letters to and from kings which were
recopied in the scribal academy as part of the instruction in learning the S umerian language"
(1980b:52). Similarly, K.R. Veenhof says:
A number of letters of kings of the Third Dynasty of Ur and some of their
successors were studied and copied in the Old Babylonian schools and
became part of the standard curriculum of that period (1986:6).
Most of these royal literary letters are products of the Ur III period. They are only
known from Old Babylonian copies; fragments of at least four copies of the letter from
Sharrum-bani to Shu-Sin are preserved.
- History
It is possible that Shu-Sin's activities did not consist of the building of a "wall" de
novo; it may have been the rebuilding of defensive fortifications started by Shulgi. Accor
ding to one letter (unfortunately, somewhat fragmentary), this wall was designed to be "26
double-hours ' : long, that is, about 170 miles ! Attempts have been made to relate this wall
to defensive lines mentioned in other texts, and even in Classical sources, and to detennine
the location of the wall, but such attempts have lacked conviction.
The wall was ultimately unsuccessful. c.J. Gadd says:
 Lesson 19 221

As for his great wall, it proved even more ineffectual than such barriers have
always been in the end. No more is heard of this vast and vain work, even
if, as seems likely, it furnished a line or a foundation for similar works in
later ages. Babylonia has no natural defences, and they were not to be pro
vided by an artificial rampart so long that it could have hardly been
effectually garrisoned ( 197 1 : 6 1 1).
It seems curious that this wall was given an Akkadian name, instead of a Sumerian
one. This may attest to the growing importance of Akkadian as the spoken language during
the Ur III period. Possibly it was chosen because its meaning would have been more-or
less understandable to Amorite speakers.
- History
The ninth year of Shu-Sin commemorates the building of the temple mentioned in this
inscription: mu _ dSaraTUmmaki-ka ba-du, "the year when the temple of Shara of Umma
was built".
Such year-dates, as laconic as they are, actually comprise one of the principal sources
of information about the history of the Ur III period.

Text 1 9a
supplementary

This text was inscribed on a votive amulet of agate.

222 Lesson 19

Sign-list and vocabulary

1 *- q EM tI-la-dBa-.!2.% Hala-Baba (PN, "the share of Baba")

=

Ur- dLamar Ur-Lamar (PN)

{10 dub-sar scribe

Commentary

2. The last sign must be a poorly-drawn IT-sign.

8. The Akkadian equivalent of b-la is zittu, from the root ziizu, "to divide". It is glossd
by the CAD as: " 1 . share of an inheritance ... ". Names of the pattern tI-la-DN are com
mon in Sumerian. Similar Akkadian names include tIA-LA-i-li and zi-i1-DIGIR, "the in
heritance of god".
9. dub means "tablet", and sar means "to write". The formation of the word dub-sar is the
same as that of zabar-dabs= sar.0 is an active participle, and dub is its incorporated object.
The meaning is thus, "he who writes a tablet". The word was borrowed into Akkadian as
tuparru (transcribed by some as tuparru). It is further discussed in Lesson 2 1 .
In historic Sumerian, sar means "to write". It i s not known what its original meaning
may have been. In its earliest attestations (para and Tell Abu Salabikh), it does not occur as
an independent verb, but only as part of the nominal compound dub-sar. dub-sar also oc
curs in the bilingual Eblaite texts, but unfortunately without an Eblaite equivalent.
 Lesson 20
This is a brick building inscription of Shu-Sin.
Sign-list and vocabulary

1J-ba-Iu5 -g Habaluge (PN)

() Adab Adab (ON)

::t. CV1f n ensi2 ensi (city governor, local ruler)

arad servant

Notes

1j-ba-Iu5-g The etymology of the name is unsure; it is discussed below.

Adab This is the name both of a city and of a kind of bird. The modern name of the city
is Bismaya. A number of third millennium (and later) texts have been found there.
The etymology of the name is unknown; it is not sure if the word is originally
Sumerian. Lexical lists, and rare syllabic spellings, most often show the initial consonant to
be Idl, but spellings with Irl and Isl also occur. Akkadian spellings also show similar
variation. Such oscillation in spelling shows that the consonant was probably not a simple
Idl, but something more complex, perhaps the IdrI phoneme discussed earlier (Lieberman
reconstructs the original name of the city and the bird as lorab/).
There is also some oscillation in the vowels of the name. The canonical spellings
show la-ai, but most earlier spellings seem to show lu-a/. The form ladabl is pre
sumably a result of vocalic assimilation: lu-a/ ) la-al. The Akkadian spellings most often
show lu-a/; that is, they reflect the Sumerian pronunciation of a period when the vocalic
,
assimilation had not yet taken place.
The name of the city and of the bird is written by what appear to be two cuneiform
signs, the ud-sign followed by the nun-sign. The situation is probably more complex,
however. According to Jacobsen,
Early occurrences show a strong tendency to combine the two later signs VD
and NUN into a ligature as if they originally formed but one single larger
sign . ... The sign is not a ligature but an original pictograph representing a
disc placed on top of a pole or stake. ... It thus becomes likely that the writing
of the city-name Adab was originally a picture of a symbol, a disc affixed to
a stake for carrying, and since that picture served also to designate the usabu
bird one may assume that the symbol represented an usabu bird and had a
picture of that bird on its disc. Symbols of this kind are well known
(1968 : 1 0 1 ).
ensi2 The function of this official has been much discussed; it changed over time.

223
224 Lesson 20

According to Hallo, the term means "titular head of a city and its dependent territories".
Jacobsen says:
The title ensik ... seems to denote specifically the ruler of a single major city
with its surrounding lands and villages, whereas both "lord" (en) and "king"
(l!!gill) imply ruler over a region with more than one important city. As for
the origins of the office, the ensik seems to have been originally the leader of
the seasonal organization of the townspeople for work on the fields:
irrigation, ploughing, and sowing (1957:384 n.7 1).
But later, during the Ur lIT period,
The top provincial civil administrators, the ensiks, became proper governors,
entirely dependent on the king, and were moved at will from one post to
another to minimize the dangers inherent in too strong local ties. Military
affairs were out of their hands entirely (1957: 155).
During the Ur ITI period, the ensi2 was the highest-ranking civilian authority; the cor
responding military officer was the akkana (often translated "military governor"). Many
of the latter were sons of the king. Occasionally, one and the same person served as both
ensi2 and akkana. The names of many ensi2s of the Ur III period have been preserved.
ensi2 was loaned into Akkadian, appearing as ii)akku and iakku. It is glossed by
the CAD as: "territorial ruler (of cities, countries, etc)".
The etymology and writing of the word are discussed below.
arad The original form of this sign was the nitab-sign (meaning "man") followed by the
kur-sign (meaning "mountain"). Presumably, the Sumerians derived some of their slaves
from foreign, mountainous areas. The shape of the arad-sign, however, tends to become
simpler, and even as early as the Ur lIT texts, the nitab-sign and the arad-sign can look quite
similar.
The Akkadian word for "slave" is wardu. This would seem to derive from the verbal
root waradu, meaning "to go down to"; this root occurs in other Semitic languages. There
fore, most scholars assume that the Sumerian word arad was borrowed from the Akkadian
wardu. A minority view, however, says that wardu has nothing to do with the verbal root
waradu, and that wardu is a borrowing from Sumerian.
A complicating factor is the fact that the reading of the sign is uncertain; some read it
as arad, and others read it as ir. The Sumerian evidence is ambiguous. According to Gelb,
"generally, the form ending in -d is younger than the form ending in -r" (1982:86). Thus,
in pre-Sargonic Lagash there occurs: (NITA1jxKUR)-ra-ni, "his slave", but in Sargonic
and Ur III texts there occurs: (NITAIJxKUR)-da-ni, "his slave".
Most scholars, however, seem to derive the form in Irl from that in Id/. For
example, Falkenstein assumes a development along the lines of: */ward-al > */urd-a/ >
*/ird-a/ > */irdl > fir!. Gelb, on the other hand, believes that there were originally two dif
ferent words for slave: a native Sumerian word, written ITI I (ARADxKUR), and also a
borrowed word, written arad (NITA1jxKUR).
There has been much recent discussion about slavery in the Ancient Near East. The
term arad is variously translated as "servant" or "slave". But as Sollberger (and others)
 Lesson 20 225

have pointed out, "The usual translation 'slave' is a misnomer because its legal implications
do not fit the Sumero-Akkadian social context" (1966: 137). In a similar vein, Gelb has
said, "Freedom is relative and the terms for 'slave' are quite ambiguous in the Ancient Near
East, as they are in the Classical World, or for that matter anywhere else" (1979a:284).
226 Lesson 20

Text 20

n I " - .----

c;:::;:;;-- 1111 r
 Lesson 20 227

Notes

Line 5: The pad-sign is here split into two parts. This is a common scribal practice with
this sign. It is presumably done for aesthetic reasons; otherwise there would be too much
empty space in the line.
Transliteration Transcription Translation

1: dSu-dZuen Susin For Shu-Sin,

2: ki-aga2-dEn-lil-h't
- - - - - -
ki.aga.a.Enlil.a(k) the beloved of Enlil,
3: lugal dEn-HI-le lugal Enlil.e the kina whom Enlil
selecte
4: ki-f!gf!2-ag4-gf!-na ki.aga.0 ag.ani.a in his loving heart,
5: in-pad i .n.pad
6: lugal-kalag-gf! lugal.kalaga the mighty king,
7: .J!gfL-
I I Unms ' ki -ma lugal. Urim.a(k) the king of Ur,
8: lugill-an-ub-da-limmuy ba lugal.anub.da.(k) limmu. bi.a(k) the king of the four
quarters,
9: digi!:-ki-g-M-f!-ni digir.ki.aga.a.ani.(r) his beloved god -
10: Ij.f!-ba-Ius-g Ijabaluge Habaluge,
1 1 : ensi2 ensi. the ensi of Adab,
12: Adabki Adab.(ak)
13: arad-da-ne arad.ani.e his servant -
14: -ki-g-M-.f!-ni e.ki.aga.a.ani. 0 his beloved temple -
15: mu-na-du mu.na.(n.)du.0 built.
Commentary

1-7. These lines are identical to lines 3-9 of Text 1 8.

9,14. Line 9 is written: digi!:-ki-g-g-f!-ni; line 14 is written: -ki-g-g-.f!-ni. In the pre
vious inscriptions, such expressions were spelled: ... -g-ni. In Text 17, for instance, lugal
ki-g-g-ni-i!: and en-ki-g-g-ni occur; in Text 19 -ki-M-g-ni occurs. Text 20 thus uses
a fuller writing. It is very difficult to understand the motivation behind such variation in
spelling.
1 1 - 1 2. Presumably a genitive phrase, ensi.Adab.(ak). As in other inscriptions, there is no
graphic expression of the genitive when the second element of a genitive phrase is a
geographical name.
1 3 . Since this is the agentive nominal phrase, there must be an ergative marker .e present.
In other inscriptions, in fact, this line is spelled: arad-da-ni-. In line 1 3, the Ii/ of the
possesive-suffix .ani has contracted into the lel of the ergative case-marker. This
contraction is indicated in the transliteration by: arad-da-ne. However, the ne-sign is the
ni-sign. Some Sumerologists would transliterate line 13 as: arad-da-ni, because they wish
228 Lesson 20

to avoid a possible prejudging of the written form. The latter view would regard trans
literations of the type "arad-da-ne" as an attempt to jiggle the script to fit our views of
Sumerian grammar. A similar instance was seen in Text 19, where ma-da-ne- occurred;
others would read this as ma-da-ni-. It is also possible to interpret writings of the type
arad-da-ni- as historical or morpho-graphemic spellings. It is such ambiguities in the wri
ting system that make it difficult to establish rules governing Sumerian phonology, espe
cially those governing vocalic contraction.
Discussion: structure
The bare-bones structure of this text is:
1 -9 For Shu-Sin benefactive
1 0-13 Habaluge agent
14 his beloved temple patient
15 built. verb
In all the inscriptions in the body of texts seen up to this point, the initial datival phrase
referred to a god or goddess, the deity for whom something was done or built. However, it
is also possible to find inscriptions recording actions performed by a subordinate of some
kind, in order to curry favor with the king. Hallo considers these to be royal inscriptions,
because there is a fair amount of latitude in his definition of royal inscriptions: inscriptions
which are dedicated either "by, or to, or on behalf of the king".
There are not many such inscriptions. Hallo lists only five building inscriptions of this
type. Curiously, they all come from the reign of Shu-Sin. king.
- Phonology
There are other cases of Idl - Isl alternation in Sumerian. Sometimes this alternation
shows up in different syllabic writings of Sumerian, sometimes it shows up in differing
Akkadian versions of loan-words or of proper names. The goddess of the scribal art, for
instance, variously appears as Nidaba or Nisaba. Some Sumerologists think that this
alternation means that Sumerian had a voiced interdental fricative, IM, which it was
difficult for the script to represent.
There are several cases of bisyllabic (or polysyllabic) Sumerian words which show
vowels of only one quality, but whose Akkadian equivalent shows vowels of two different
qualities. For example, there is a tree usually spelled za-ba-Iam in Sumerian. The word
appears in Akkadian as supalu. Most probably, this is a pre-Sumerian substrate word,
which passed into Sumerian. It then passed into Akkadian, presumably through Sumerian
(or conceivably by a different route). The Akkadian preserves the older vocalization; the
Sumerian form shows that at some time Sumerian underwent a rather extensive process of
vocalic assimilation, a process of which only traces can be seen. Assuming that Akkadian
borrowed the word from Sumerian, it must have borrowed the word before the Sumerian
vocalic assimilation had taken place. This same process was seen in Lesson 1 1, where the
word zabar was discussed; the pronunciation Izabarl is the result of vocalic assimilation
from some earlier form of the word like Isibar/; Akkadian again preserves the more
 Lesson 20 229

original vocalization.
- ensi2
There is no unanimity among scholars about the etymology and the writing of the term
ensi2. (The latest summary is in Dunham 1986:5 1-52.)
The word probably had a Ik/-Auslaut. When it is followed by a vocalic ending of
some kind, a Ikl usually appears; note also the Akkadian ii'akku. It has frequently been
proposed that the Ikl is the genitive marker; the word may originally have been a genitive
phrase, en. si. (k), "the lord of si". Unfortunately, it is not sure what si means here.
Jacobsen thinks that si means "arable land"; the title would originally have meant "manager
of the arable lands". But it has also been argued that the word has a pre-Sumerian substrate
etymology, and that the interpretation "lord of si" is a Sumerian folk-etymology.
Some Sumerologists now transliterate this word as enS -si. This interpretation is
partially based on the understanding of the word as a genitive construction, even though the
spelling with en S (instead of simple en) is weird.
The problem of understanding the writing is complicated by the fact that although
normally written PA-TE-SI, the word is also occasionally written PA-SI, and even just PA;
other bizarre spellings are attested. These writings make it difficult to understand how the
cuneiform signs are meant to refl ect the language (should they be regarded as
abbreviations?). Such writings cause numerous problems in transliteration.
-Genitive
One of the pieces of evidence which indicates that the final Ikl of the genitive
morpheme was actually pronounced in spoken Sumerian is the fact that it occasionally turns
up in words loaned into Akkadian. For instance, the Sumerian god Amar-Utu, "young bull
of Utu", appears in Akkadian as Marduk, presumably from amar.utu.(k) If the Sumerian
ensi2 is indeed a genitive phrase, then the Akkadian ii)akku also shows the genitive mar
ker Ik/. It has, however, also been argued that "Marduk" is not of Sumerian or of Ak
kadian origin, and has nothing to do with "Amar-Utu".
- Proper names
The personal name "Habaluge" occurring in this inscription is presumably of
Sumerian origin, although the exact etymology is not known. When a name is of uncertain
etymology, it is often difficult to determine the precise reading of the signs. For example,
the third sign in this name can be read Iu S ' lul, lab, nar, etc. The fourth sign is also gra
phically ambiguous; several different signs share approximately this same shape. For
example, the -sign and the ke4 -sign, even though distinct signs, look very similar in sev
eral of these royal inscriptions. And if it is the ke4 -sign, it has several different readings:
ke4' , lil, etc.
One way such ambiguities can be resolved is to identify different spellings of the same
name. For example, this particular name is not uncommon in Sumerian texts. It is most
commonly transliterated as: Ij-ba-lu S -ke4' with the proviso that the reading of the third
and fourth signs is not certain. There are also spellings where the third sign appears as lu,
230 Lesson 20

and not as the lus -Iul-Iab-nar-sign: lj-ba-Iu-g and lj-ba-Iu-g 18 . This shows that the
third sign is probably to be understood as Ilu/, and so to be read as lu, and not as lul, lab,
or nar. The fourth sign is spelled as , g, and g 18 : lj-ba-Ius -' lj-ba-Iu-, and lj
ba-Iu-g 1 8. This shows that the last sign is a g/ke4 -sign, and not the graphically similar
sign; it also shows that the reading was probably Ige/. Therefore, the most likely reading
of the name in Text 20 is: lj-ba-Ius-g.
The latest discussion of this name is by Steinkeller ( 1984:9); he reads the third sign as
lugx' that is, with a Ig/-Auslaut. He believes that the meaning of the name "cannot be
gauged with confidence", but it is undoubtedly a Sumerian verbal form with the modal
prefix b; it may mean "May-he-pasture/take-care-of '.
 Lesson 2 1
The three texts in this Lesson do not offer very much new in the way of grammar, but
they illustrate common types of royal inscriptions. The first is a weight of Shu-Sin. The
next two are seals of rbbi-Sin, the son and successor of Shu-Sin; he was the last ruler of
the Ur m Dynasty.

Sign-list and vocabulary

W "5"
ma-na "mina" (measure of weight, about 505 grams)

&i-oa true. correct; standard. certified

Notes

5 When Sumerian numerals are used strictly for counting, they are normally transliterated
by Arabic numerals. The word for "five" was iA.

ma-Da hs etymology and value are discussed below.

gi-na This is a loan from Akkadian klnu.

Text 2 1 a

231
232 Lesson 21

Transliteration Transcription Translation

1 : ma-na gi-na 5 mana gina 5 standard minas.

2: dSu- dZuen Susin Shu-Sin,
3: lugal-kalag-g lugal.kalaga the mighty king,
' ki-ma
I !wI- Unms
4: ...! lugal.Urim.a(k) the king of Ur,
5: lugru-an-ub-da-limmuTba lugal.anub.da.(k) limmu. bLa(k) the king of the four
quarters.
Commentary
1 . In simple enumerations, the numeral is followed by a singular, not a plural. In many
languages, the noun appearing after a numeral assumes special forms, in number or in case.
Sumerian shows no outward evidence of this; in general, agglutinative languages show no
special forms after numerals.
2-5. There is no verb form.

Discussion: weights
This is a typical weight inscription. Unlike the weight inscription seen in Lesson 10,
this one actually gives the weight. It is difficult to say exactly what gi-na means in such
contexts; it is usually understood as "standard", or perhaps "certified". One of the more
important functions of Mesopotamian rulers (throughout all periods of history) was the
regulation of the system of weights and measures, but not much is known about how such
weights were actually managed by the crown. Gadd points out that Shulgi "rearranged the
calendar, set up a bureau of standards, and issued accurate weights which were preserved
and imitated to the latest days of Babylonian history" (197 1 :618). In the prologue to Ur
Nammu's Law Code, there is a section referring to the "standardization" of the mina; this is
usually interpreted as referring to some kind of reform of the royal weight system. Re
cently, Irving Finkel has published a text dated to Amar-Sin's first year, which is a "receipt
for two differing sets of weights":
The implication of the text ... is that an official issue of correct weights was
made at the beginning of Amar-Sin's reign, and that this document reflects a
deliberate attempt to ensure that government offices were using uniform
weights (1987 : 192- 193).
- Metrology
All the Classical Semitic languages except Ethiopic have a verbal root *mnw/y,
meaning "to count, to reckon". Therefore, Sumerian ma-na is probably a loan from the
Akkadian manu (which is also the ultimate source for the English word "mina").
The value of the ma-na and the manu varied to some degree from time to time and
from place to place (Powell refers to "a multiplicity of standards which defies reduction to
one or more 'common' norms"). In Sumer proper, the most common value of the ma-na
was about 505 grams. In Mesopotamia, the manu was the same. But in most of Syria, the
 Lesson 21 233

manu was a little less, from 470 to 480 grams. At Ebla, it was also about 470 grams. (The
theoretical value of weights is determined simply by averaging out the actual weights of
weights inscribed with their values.) Since this particular weight is a .5. ma-na weight, it
should weigh about 2525 grams. It actually weighs a little less, 25 1 1 grams.
The Sumerian ma-na was divided into 60 gin; the Akkadian manu into 60 iqlu. 60
ma-na formed a gun; 60 manu formed a biltu:
= 60 ma-na 1 biltu = 60 manu
= 60 gin 1 manu = 60 iqlu

- Numbers
The pronunciation of the numeral for "five" as fia/ is known from mathematical
cuneiform texts, where it is occasionally spelled out. The Ebla school-text mentioned in
Lesson 10, which spells out the Sumerian numerals from one to ten, simply gives i for "5".
Unfortunately, not enough is known yet about the nature of the Eblaite syllabary to say
what values the i-sign could have had at Ebla; however, there is some evidence to show that
one of its values was /ya/. Pettinato, in fact (198 1 : 143) reads the sign on TM.75.G.2198
as i, but this is perhaps somewhat adventuresome.
It is also significant that the i-sign is composed of five strokes: ;::i= .
>--

Sign-list and vocabulary

> Nanibgal Nanibgal (DN)

%--

3=* I-bi -dZuen Ibbi-Sin (PN)

B=J 8=J Da-da Dada (PN)
*
Ur-dNanibgal
ill *- Ur-Nanibgal (PN)

zu your

Notes

I-bi-dZuen The name is good Akkadian, meaning "Sin has called"; ibbi is the preterite
from nabu, "to call".
D a-da The etymology of the name is unsure; it is discussed further below.
Nanibgal Very little is known about this deity. The reading of the name is somewhat un
certain, as is the etymology. The cuneiform character appears to be the an-sign followed by
the nisaba-sign. Therefore, the name is sometimes transliterated as AN.NISABA, or as
234 Lesson 21

DIGIR.NISABA (Nanibgal is not infrequently mentioned alongside Nisaba). In the writing

in Text 2 1 b, the determinative and the an-component of the sign are written on top of each
other, instead of following each other. This is for the sake of graphic symmetry; other
times the two components are written after each other.

Text 2 1 b

Notes

The name of Dada's father, occurring in line 1 1 of this text, illustrates the difficulties of
working from autographs. The autograph seems to show the name as Ur-t!gS . t!gS (or
a6) means "good". The name would then be a variant of Ur-ags-g, a relatively common
personal name meaning "the good man". However, the photograph of the text is more
ambiguous. The sign in question is actually somewhat damaged, and the remaining traces
can be made to fit either ags or dNanibgill.
The reason for preferring the reading dNanibgal over t!gS is because other seal im
pressions have been preserved of this same individual, Dada, in which the sign for the
name of his father is more distinct, and in these the sign is clearly Nanibgal.
One cannot always accept a modern-day editor's transliteration of a text. An autograph
carries more evidential value, but even then cannot always be accepted at face value. This is
especially true for autographs which were drawn when knowledge of Sumerian was weak.
Better than a transliteration or an autograph is a photograph. However, for many published
texts no photographs are available, and for others the photographs are reproduced in such
poor quality that they are almost useless. For any significant passage, there is no alternative
to a close examination of the original cuneiform document.
 Lesson 21 235

Transliteration Transcription Translation

1 : d!-bi- Ibbisin Oh Ibbi-Sin,
2: dZuen
3: digir-kalam-ma digir.kalam.a(k) god of the land,
4: lugill-kalag- lugal.kalaga mighty king,
5ki-ma
5: -IYgIDI- Unm lugal. Urim.a(k) king of Ur,
6: lugal-an-ub- lugal.anub.da.(k) limmu.bi.a(k) king of the four quarters -
7: da-limmu2-ba
8: Da-da Dada Dada,
9: ensi2 ensi the ensi of Nippur,
10: Nibruki Nibru.(k)
1 1 : dumu Ur-dNanibgill dumu. Urnanibgal the son of Ur-Nanibgal
12: ensi2 ensi the ensi of Nippur -
13: NibrukCka Nibru.k.a(k)
14: arad-zu arad.zu is your servant.
Commentary
1 . It is usually assumed that in seal inscriptions the initial nominal phrase contains the
name of the king in a vocative. The vocative normally has no formal marking in Sumerian.
Another example occurs in Text 22. There are a few cases where the vocative is marked by
.e. This is presumably an extension in use of the locative-terminative.
The essence of this seal is: "Oh Ibbi-Sin, Dada is your servant".
1 3 . A double genitive is contained here:
Dada,
ensi.Nibru.(k)
dumu. rUrnanibgal l
Lensi.Nibru..a(k)
14. zu is the second person possessive-suffix, "your". It is difficult to say whether this is
a nominal sentence without the copula ("Dada is your servant"), or an appositive ("Dada,
your servant").
Discussion: possessive-suffixes
The possessive-suffixes for the singular are:
first person - g!! lQ (mu, g!! IQ)
second - zu
third animate - .!!-ni
inanimate - bi
236 Lesson 21

The initial consonant of the marker for the firstperson singular is the velar nasal
discussed under Phonology. conventionally transliterated by Sumeroiogists as g. There
fore, gylO is the most up-ta-date reading of this sign in this usage. However, it is also
transliterated as gylO' and (especially in older works) as mu.

- Proper names

The name Da-da occurs very frequently in cuneiform texts. It is not very easy to de
termine the etymology of a name of such a simple structure, what appears to be a redup
licated CV syllable. The name has variously been considered to be Sumerian, Akkadian, or
"other", In an article on "Ethnicity and Onomastics in Sargonic Mesopotamia" (1982),
Benjamin Foster divided personal names into four groups: Sumerian; Akkadian; Redup
licated; Unsure. He purposely omitted the name -IDt from discussion, because of the
difficulty in detennining its etymology.
Such reduplicated personal names are sometimes referred to by the Gennan tenn
"Lallnamen" or "LallwOrter" ; another example is Du-du.

-Seals

Seals were used by officials in Mesopotamia (and elsewhere) to stamp their "seal of
approval" upon documents of all kinds. The act of sealing could perfonn several functions,
such as acknowledgment, authorization, guarantee, etc. Mesopotamian seals usually
consist of two components: a pictorial scene of some kind, and a short inscription. Many
seals have only a pictorial scene, and lack an inscription. There are only a few seals which
have just an inscription, but lack a pictorial scene. The pictorial representation on the seal
frequently has a mythological significance, as in the example given below. Or, it may be a
"presentation scene", usually thought to represent the possessor of the seal paying homage
to his ruler, the king. The following is a picture of a (non-royal) seal, from the Old
Akkadian period. (This partiCUlar seal has often been reproduced.) The inscription reads:
M-d..!! mm-sar, "Adda the scribe". The photograph is actually of the impression of the seal,
not the seal itself. This is because the text on seals is inscribed in reverse (that is, mirror
imaged), so that when impressed, the text comes out in the right direction.

I II
'f ,

)j\; I
'

A
, \ '-'-'-'-
-
 Lesson 21 237

It is not easy to detennine what language such a short inscription is written in. It may
be perfectly good Sumerian. However, according to the criteria used by art historians, the
seal in the photograph dates to the Old Akkadian period. Therefore, the language is
probably Akkadian, and dub-sar should be read as a logogram for tuparru.
- Scribes
Scribes occupied a central role in all of Mesopotamian civilization. Many modem
scholars have written about them, but there is still much that is not known; for example, to
what degree was their job hereditary; what was the extent of their influence at the royal
court; did they have non-scribal work at the court or elsewhere, etc. Writing on the "social
position of Neo-Babylonian scribes", Muhammad Dandameyev has said:
Our infonnation about the social position of the Mesopotamian scribe, his
activity as bureaucrat and in the service of the community for recording of
contracts is very scanty. We have no direct data on the economic situation
and the social origin of scribes. We also do not know if the scribal
profession was the chief source of income or if the scribes were busy with
their craft along with handicraft, tilling ofland and so on ( 1982:35).
C. B . F. Walker points out that
The scribes, like any craftsmen, had to undergo training, and having
completed their training and become entitled to call themselves dubsar
"scribe", they were members of a privileged elite who might look with
contempt on their fellow citizens (1987:33).
Specifically discussing seal-practice in the Ur III period, Steinkeller says:
The tenn dub-sar, apart from its basic meaning "scribe" is an honorific title
which merely indicates the graduation of the individual in question from a
scribal school. ... It is tempting to speculate that the "dub-sar seal" was a
kind of "diploma", which may have been presented to a graduate of a scribal
school at the conclusion of his studies. The possession of such a seal would
have constituted proof that its owner was eligible and entitled to be employed
in the state or temple administrative apparatus or to sell his services to private
individuals ( 1977:47-48).
In a similar vein, Veenhof mentions the role of scribes
in the administration and their position in society, which may range from
that of a simple clerk or a paid letter writer on the market to that of a chief
accountant or secretary of a chancery or king (1986:2 1 ).
The rather automatic translation of dub-sar as "scribe" paints a rather simplistic picture.
Michalowski says that "In Ur III times dub-sar was a general term for low and middle level
bureaucrats" (1987:62).
And to quote Walker again,
[Most scribes,] after all their technical training, spent their lives writing lists
of deliveries of sheep or issues of barley rations and occasionally taking a
letter by dictation. The more successful scribes would end up as senior
administrators in the state bureaucracy, but most of their colleagues would
238 Lesson 21

have been happy simply with their status as educated men and the
knowledge that their training guaranteed them employment ( 1987:39).
In a thought-provoking article about what we don 't know about Ur III society, Soll-
berger asks:
We know roughly what the professional scribe's jobs consisted of, but how
did he work, and where did he work, and how did he make himself known
as a professional scribe and his services available? And there is of course
the nagging question which is usually politely glossed over: where did the
scribes get the enormous amount of clay they needed? Were there clay
stationers? Did one have to buy clay or did one just go to the canal bank and
help oneself ... ? ( 1972: 1 88).
- Engravers
When dealing with "monumental" inscriptions, it is necessary to distinguish between
"scribe" and "engravers" (or, "lapidaries"). The latter were the persons who actually
chiselled the inscriptions into the stone. They were not always literate, but simply copied a
design or plan, which may have been drawn onto the stone. Presumably, the engraver
worked under the supervision of a scribe. In other cases, the scribe and the engraver may
have been the same person.
The standard word for "engraver" was zadim. This word apparently derives from za,
"stone", and dim, "to fashion"; dim is an active participle, and za is an incorporated direct
object.
In the case of seals, the situation was probably a little more complicated. The
inscription and the pictorial scene were sometimes engraved by different individuals. W.
G. Lambert (discussing seals of the Cassite period) asks:
A basic question which needs answering is, who carved these inscriptions?
Did one man carve both glyptic and inscription, or were separate craftsmen
employed for the artistic and scribal parts? In some cases it is clear that the
glyptic was carved first, because not enough room was left for the
inscription, so that the last line had to be spread out among the glyptic. But
in other cases where the inscription covers virtually the whole area, and the
glyptic is reduced to a row of insects for example, then one may suspect that
the inscription was carved first and the glyptic was a second thought, serving
merely in fugam vacui. ... One may wonder if two quite separate guilds of
craftsmen were in existence, and such a division of labour seems very
probable in the contemporary boundary stones. ... On general grounds too
such a differentiation is likely, since the artist and the scribe needed very
different training. Yet one need not suppose that this demarcation was al
ways completely enforced ( 1975:220).
- Seals
Many seals from the Ancient Near East have been preserved. Even more common
than the seals themselves are seal impressions, that is, the impression of a seal upon a
 Lesson 21 239

cuneifonn document. Text 21b is a seal-impression found on a record of official ap

pointments of individuals to sundry governmental offices.
Seal impressions are necessarily rather small, being squeezed onto a small seal. This
accounts for some of the odd division of names and epithets put onto more than one line.
This small size often makes it difficult to read autographs or photographs of seal
impressions, even if reproduced full-size. Occasionally, the script used on the seals is
archaicizing, compounding the problems ofreading.
When publishing editions of cuneifonn texts, Sumerologists do not usually present a
drawing of a complete seal as it appears on a document; they will usually only reproduce
the impression left by the inscription. This is primarily because of the mechanical effort it
takes to adequately reproduce (and even just to describe) the scene carved on the seal, and
linguists are not artists or art-historians. Sometimes the seal impressions will be briefly
described, and occasionally published in a separate volume, distinct from the texts
themselves.
The practice of only reproducing the inscription, and not the pictorial scene, is
unfortunate. While linguists may only be interested in the inscription, art historians,
anthropologists, and historians, among others, are just as interested in the scene itself. Seal
impressions are also important for scholars studying groups of documents, and archival
relationships.
- arad-zu seals
Texts 21b and 21c may be called royal seal-impressions, again following Hallo's
definition of "royal" as: "by, or to, or on behalf of the king". Many non-royal seals and
seal-impressions (such as that in the photo above) have also been preserved.
There are two principal types of Ur III seals. Text 21 b is an "arad-zu" seal. This type,
which is very common, has a specific structure. First, there is the name of the ruling
monarch. This is a nominal phrase, in the vocative. Second, there is the name of an
official, with various epithets or filial relationships as appositives. Third, the tenn "your
servant" concludes the seal.
The usual interpretation of arad-zu seals is that an official had it cut, out of homage or
respect for the king. However, it has been speculated by Richard Zettler that "the flow of
these seals was from king to official and not from official to king" (1977:33).
A few seal-inscriptions have been preserved where the last line reads arad-ni, not
arad-zu. Presumably arad-ni is for arad.ani, "his servant". (This is another instance of the
overhanging vowel problem.) Not enough is known about such seals to characterize them
differently from the arad-zu seals.
Although arad-zu seals have traditionally been classified as "Sumerian", with a
Sumerian inscription, it has been proposed to read seals of this type from the Akkadian
period as Akkadian. That is, arad-zu is to be read as ARAD-su, for warassu, "his servant".
(This is the usual Akkadian fonn, resulting from the regular assimilation of */dU ) Iss/:
*/warad-ul ) Iwarassu/.) The rest of the text would then be understood as logograms, to
be read in Akkadian.
The reason for this possibility is because of a seal where the last word is written arad-
240 Lesson 21

za, presumably ARAD-sa, for Iwarassa/, "her servant", and another seal written geme2-za,
,
possibly GEME2-sa, for lamassa/, "her servant (fern) '.
While this may be true for seals of the Akkadian period, it is hard to say whether it
might be true for Ur III (and other) seals. There is really no evidence to decide one way or
the other, and so the question is still open. Barring explicit evidence to the contrary, it is
probably best to assume that for the Ur III period, at least, the text is Sumerian.
- Seals
Both Text 2 l b and 2 l c may be called functional, in the sense discussed in Lesson 4.
There are also a certain number of votive seals. These are known primarily from the seals
themselves; only a few of the actual impressions are preserved. Text 22, below, is such a
votive cylinder seal.

Sign-list and vocabulary

GJ:::=f 1!f Sag-dNanna-zu Sagnannazu (PN)

sa (kind of priest)

r:8 ba to give as a gift

Notes

Sag-dNanna-zu sag means "head", but can also have the meaning "slave". zu is a verbal
root meaning "to know". Names of this type are usually interpreted as some kind of
reduced relative or participial clause, "the slave who knows Nanna", or "the slave whom
Nanna knows".
saga This priest was high up in the temple hierarchy, although very little is known of his
priestly duties. He seems to have been mostly concerned with running the administrative
side ofthe temple. While the conventional translation is "sang-priest", Gelb points out that
"In his capacity as the head of a household, the word san may be interpreted as the chief
administrator of a temple household" (1979b: 1 6); Snell translates the title as "economic
director of a temple". The same cuneiform sign, in fact, can be read as ita5 ' "to count"
(Akkadian manu); Aage Westenholz, among others, says that the sign "depicts an abacus or
a countingboard" ( 1985:296).
 Lesson 21 24 1

Text 2 1 c

Transliteration Transcription Translation

1: dI-bi- dZuen Ibbisin Ibbi-Sin,

2: lugill -kalag-ru! lugal.kalaga the mighty king,
3: ...!
I !g-
I Unm' Ski-ma lugal.Urim.a(k) the king of Ur,
4: lugal-an-ub-da-limmu T lugal.anub.da.(k) limmu.bi.ak.e the king of the four
- oa-=ICe4 - quarters -
5: Sag-dNanna-zu Sagnannazu to Sagnannazu,
6: salli!_dEn-lil-hi
- -
- -
saga.Enlil.a(k) the sangf!.-priest of Enlil,
7: arad-da-ni-i!: arad.ani.r his servant -
8: in-na-ba Lna.(n.)ba gave (this seal).
Commentary
8. The verb form is almost always written this way. The in-sign contains the conjugation
prefix .i and the initial I nl of the datival dimensional-prefix.
Discussion: arad-da-ni-ir seals
This type of seal is known as an arad-da-ni-ir seal, or as an in-na-ba seal. Its most
common structure is: First, the name of the king, with the ergative case-marker; second, the
name of an official, with various epithets; third, the appositive arad.ani, with the dative
case-marker; fourth, the verb form in-na-ba. This results in the datival noun phrase
following the subject, and immediately preceding the verb. In most of the texts seen earlier,
the datival noun phrase occurred at the beginning of the text.
The understood direct object in this type of seal is the seal itself. The king gave such
seals to his officials (and family members), presumably as a reward for some kind of
service. For this reason, they are sometimes referred to as "presentation seals". It is also
possible that the king gave the seal to an official upon his appointment to an office. Fewer
such seals are preserved than arad-zu seals.
arad-zu seals are found under all five kings of the Ur III Dynasty. arad-da-ni-ir seals,
on the other hand, are only attested for the reigns of Shu-Sin and Ibbi-Sin. This may (or
may not) be due to accidents of preservation. It is also possible that there was some kind of
change in administrative practice.
242 Lesson 21

- History
Shu-Sin apparently died a natural death. He was succeeded by his son, Ibbi-Sin.
Early in the latter's reign, the eastern territories under the control of Ur broke away, then
other parts of the empire began to fall away. For most of his reign, Ibbi-Sin's control
extended no further than the city of Ur itself. The economy collapsed, and a vicious
inflationary spiral ensued. Very little is known of the details of the twenty-four or twenty
five years of Ibbi-Sin's "reign". Jacobsen has said:
How an empire like that of the Third Dynasty of Ur - to judge by our
sources the most efficiently organized structure of its kind before Assyrian
times - could so quickly and so completely collapse without pressure from
any enemy state or states of comparable magnitude is really quite puzzling
(1953: 173).
Jacobsen wrote this over thirty years ago, but his puzzlement still largely stands;
Steinkeller says that "the phenomenal rise of this empire was matched only by the sudden
ness and completeness of its demise; in less than a century after its creation, no trace of it
remained" ( 1987b: 19).
Finally, there came invasions by the Amorites, against whom Shu-Sin had built the
wall mentioned in Text 19, and the Elamites. The Elamites, aided by a somewhat obscure
group of people from the Zagros mountains known as the "Su" or "Sua" (recently
identified by Steinkeller [1988] with Shimaski in Iran) sacked Ur, then withdrew back to
Elam, carrying Ibbi-Sin back with them; he died in Anshan. Gadd says: "Ibbi-Sin became
the typical figure of an ill-starred king, remembered only for his captivity and death in a
strange land" (197 1 :6 17).
During their sack, the Elamites destroyed every temple standing in Ur, and all of its
fortifications. One of the most well-known Sumerian literary compositions is a long poem
entitled "The Lament over the Destruction of Ur", which bemoans its destruction.
This was not the only time that Ur was sacked; Samsu-Iluna of Babylon also levelled
the city, in 1740 BC. The year-date for the eleventh year of his rule is: "The year in which,
at the pleasure of Anu and Enlil, he destroyed the walls of Ur and Uruk". Woolley
describes the destruction:
The ruins bear eloquent testimony to the thoroughness of that destruction.
The fortifications were dismantled - this indeed one might expect; every
temple that we found had been plundered, cast down, and burned; every
house had been consumed with fire; the whole of the great city ceased to
exist (1982:214).
However, the city was rebuilt almost as often as it was sacked. Most kings of the
Isin-Larsa period rebuilt old temples and built new ones. Such construction took place
right through the Neo-Babylonian period. Although Ur never regained the political
importance it enjoyed under the Ur III Dynasty, there were times when it must still have
been an imposing city.
Ur was occupied - at least, to some degree - into the Persian period. It began to
completely fade out about the fourth century BC, through the effects of changing trade
 Lesson 21 243

patterns. the shifting of the course of the Euphrates and the concomitant loss of agriculture,
etc.

Text 2 1 d
supplementary

Notes

Line 8, In Text 21a, gi-na (a borrowing from Akkadian) was used as an adjective.
However. it can also be used as a verb: mU-!li!-gi-in. It is usually translated as "he stan
dardized". or perhaps here "he certified". As discussed above, the exact significance of the
tenn is not clear.
The actual weight of this weight is 2478 grams.
 Lesson 22
This is a votive cylinder seal. made of limestone. dedicated to the tife of Shulgi.

Sign-list and vocabulary

<ID r:t>l <:) Jll:J n Mes-Iam-ta--! Meslamtaea (DN, mase)

8":iiij Ki-Iul-Ia Kilula (PN)

L!:::l r8 J.!r-l1lI-? Urba-x (PN)

--<:7-(> LagaS <SIR.BUR.LA) Lagash (ON)

,iliiJ arm, strength

245
246 Lesson 22

'(i>---7 ff gg-za throne

lJ>--t r gg-za-hi (kind of official)

qro kiib cylinder-seal

d T>-- '@ getug3 ear; intelligence

JJT" nig (n1) thing

r hi to hold, to lift, to carry

::m ag5 (g6) to be good, pleasant, nice

il

Notes

Mes-lam-ta-e-a God of the Netherworld, apparently the same as Nergal. Mes-lam (also
transliterated as MM-lam) is the name of the temple of Nergal in Kutha. Its meaning is
unknown (although gimes is well known as a type of tree). ta is the ablative case-marker.
r:) )I:::j is a verbal root, meaning "to go out". is the nominalizer. Thus, the name means
something like "he who goes out of the Meslam-temple".
Meslamtaea himself does not occur frequently in texts, but Nergal was widely
worshipped in Mesopotamia. He shared the rule of the Netherworld with Queen
Ereshkigal.
Ki-lul-la This not uncommon name is presumably Sumerian, although the etymology is
unsure. It is variously spelled as: Ki-lul-la, Ki-lul-hi, and Ki-lul-hl-.
Ur-ba-? The reading of the third sign is uncertain. Both the autograph and the photo
show a sign which looks closely like the last sign in the inscription, which must be bi. The
reading bi for this sign is also the reading preferred by Gadd, who says that upon a "fresh
examination" of the seal, "the engraver certainly traced, and doubtless intended, the same
sign as in the last place of the whole inscription", that is, bi. However, there seem to be no
parallels to a putative name Ur-ba-bi. (Gadd speculates that Ba-bi may be a variant of the
name of the goddess seen in Lesson 1 1 , spelled as dBa-.!:2% , which earlier Sumerologists
read as dBa-.!1.)
Most Sumerologists have interpreted the sign as a poorly drawn -sign (if not a
simple mistake). The g-sign here would be read as gar. Ba-gar is the name of a well
attested temple of Ningirsu at Lagash (the etymology of Ba-gar is unsure). The personal
name would then mean, "the man of (the temple) Bagara". This name is attested elsewhere.
Also, other names composed of Ur and the name of a temple are attested.
 Lesson 22 24 7

Lagd "Lagash" is used in two senses in Sumerian (and in English). Strictly speaking, it
refers to a city proper (whose modern name is el-Hibbe). But it can also be used to refer to
the territory controlled from the city of Lagash. This larger entity is sometimes referred to
as "Lagash-state", as opposed to "Lagash-city". This included Lagash-city; the adjacent
Girsu (whose modern Arabic name is Telloh, meaning "mound of the tablet"), and Nina
(modern Surghul). Because of this ambiguity, the term "Lagash" is sometimes used when
referring to Girsu.
This was the site of the first important excavation of a Sumerian tell, begun by the
French in 1 877. Thousands of tablets were found, including a number of royal inscriptions
from the First Dynasty of Lagash, and from the time of Gudea. These are one of the prime
sources of Sumerian for the period.
The etymology of the name is unknown. It is not known how the three signs SIR
BUR-LA came to represent the name of this city. One would guess that the la-component
is some kind of phonetic complement. The name is, in fact, occasionally written LA-BUR
SIR.
It This originally meant " arm", and then "strength". In this latter sense, its Akkadian
equivalent is eml1q!!, glossed by the CAD as: " 1 . strength (in physical sense as localized in
the arm), . .. ". J!-zid-da is something like "the effective arm". It can be approximately
rendered in English as "right arm" or "right-hand-man".
In Text 17, the expression digi!:-zid occurred, translated as "the effective god". There,
zid appeared without the nominalizer -. But here, the same adjective occurs with the
nominalizer -: zid-da.
gu-za-llt g!!-za means "throne". hi is a verbal root meaning "to hold, to lift, to carry".
Here, hi is an active participle, and g!!-za is its incorporated object: "he who holds the
throne", or "throne-holder".
gg-za-la was borrowed into Akkadian as guzalU, explained by the CAD as: "an offi
cial, lit. chair-bearer, originally a servant carrying a chair after his master".
gg-za is thought to be the ultimate source of the word for "chair, throne" appearing in
many Semitic languages: Akkadian kussu, Hebrew kisse, Arabic kursTy, etc. But since the
pattern of the Sumerian form is a little strange, it has more recently been speculated that gg
za is a borrowing from Akkadian. This situation illustrates the difficulty of evaluating the
evidence of loan words; it is not always easy to establish the direction of a borrowing.
kBib It has been speculated that this is a pre-Sumerian substrate word.
getug3 This can be spelled in a variety of ways. The simplest was originally the picture
of a donkey's ear ( <T>-), transliterated as getug. It can also be written with the signs
and illg functioning as phonetic complements; the combination of these three individual
signs can appear in varying orders:
ge-tuggetug ( getug2)
=

gegetugtug ( getug3 )
=

Although it is clear (now) that ge and illg in this particular use are phonetic
248 Lesson 22
complements, they are not usually transliterated as such. Since the simplest (at least,
phonetically) reading of the getug-sign is m, Sumerologists often refer to this sign as the
"m-sign", and transliterate this word as: ge-tUg-m, or (as here): ge-ill-tug.
In its concrete meaning as "ear", the Akkadian equivalent of getug is uznu. However,
it is also equated with basTsu, glossed by the CAD as: "(1) aperture of the ear, ear, (2) (fa
culty of) hearing, (3) understanding". Jacobsen says that "the Mesopotarnians believed the
ear, not the brain, to be the seat of intelligence".
ags Variously transliterated as: ag5 ' a6 ' sag9 ' sa6 ' sig6 , etc. The most common Akka
dian equivalent is damliqy. The verbal adjective damqy is glossed by the CAD as: " 1 .
good, fine, pleasant, 2. beautiful, handsome, 3 . of good family, well-to-do, 4 . expert,
well-trained, 5. of good quality, in good condition, 6. gracious, favorable, 7. propitious,
8. effective, 9. canonical".
 Lesson 22 249

Text 22

* J:1>J C} E }F
j3 H .El
-+ Efl tIr ni l I
, (
t ::\r ,

*tm=f ,\ tit) 11 " E!Jll>

ElW <Q B El
W tII [t>--( ff r
ll::f UII I
\A,-r 4 rEI-
E!tfiP> :,t!.igl fJ\,!t
:.::. : : *, fr>
-=

* JIIUt, ,::><J
u

l
250 Lesson 22
Notes

Line 9. Assyriologists use a superscript exclamation mark ( ! ) to indicate that the cuneiform
sign so marked in transliteration is written in a deviant or aberrant fashion on the tablet.
For example, the proper name in this text might be transliterated as: Ur-ba-gar ! . This
transliteration means that in the opinion of the modern editor, the sign is a poorly drawn
gar-sign.
Another convention is to employ the superscript exclamation point to mark the correct
value in transliteration (that is, what the editor thinks it should be), and to follow this with
"what is actually written", in caps within parentheses. Thus, a transliteration such as: Ur
ba-gar! (BI) would mean that in the opinion of the editor, the sign drawn on the tablet is
bi, but the editor believes that this bi-sign is a mistake for gar - the scribe made an error.
As can be imagined, these conventions can be misleading or confusing. They illustrate
the importance of working directly from the texts, from photographs when possible, or
from autographs, and not just from transliterations.
Line 10. Because of damage to the cylinder seal, this line is somewhat difficult to read. Of
the getug3-sign, the ge-component o;=f) is clear. The ID (or getug)-component J>--) is
virtually completely effaced. Of the !!1g-component ( j' ), only the bottom shows. In
addition, the following nig-sign ( ilY") is scrunched up.
Transliteration Transcription Translation
1 : des-lam-ta--g eslamtaea For eslamtaea,
2: lugal -zid-da lugal a.zid.a the king, the right-hand man
of
3: Lawki -ke4 Laga.(a)k.e Lagash -
4: nam-.ti-il nam.til for the life of
5: dSul-gi nitab-kalag-gg Sulgi nitab.kalaga Shulgi, the mighty man,
6: lugill-Urim5ki-ma-ka-e
- -- -- - - lugal.Urim.ak.a(k).e the king of Ur -
7: Ki-lul-la g!!-za-la Kilula guzala Kilula, the guzala-official,
8: dumu Ur-ba-?-ke4 dumu Urba? ak.e the son of Ur-ba? -
9: mu-na-dim kiib-ba mu.na.(n dim fashioned (this). The name
C
kiib.bi.a k) of this cylinder seal is:
10: lugill-gu l O Retug3 -nig-ag 5- lugal.gu .getug.nig. ag. "0 my king, let me keel? him
gg-ICa-ne ak.am.e alive at his ear of favor .
1 1 : gg-an-.ti-il ga.(L)n.til
12: mu-bi mu.bi
Commentary

3. Laga.ak.e, written Lagaki-ke4 ' The followers of the Falkenstein school would read
the first sign as Lagaa.
The .e at the end is the marker of the locative-terminative case. In all the previous
 Lesson 22 25 1

instances of indirect objects and benefactives, the nominal phrase used the dative case,
marked in .ra. Occasionally, however, the locative-terminative case, marked in .e, is used
instead. This reflects a further spread of the usage of the locative-terminative; such a use of
the locative-terminative for the dative does not appear in earlier Sumerian.
The combination of the Ikl of the genitive marker with the lel of the locative
terminative is written with the ke4 -sign, just as is the combination of the Ikl of the genitive
marker with the I el of the ergative. (This ambiguity of the ke4 -sign can lead to confusion.)
4. Here and in line 1 1 , the root for "to live" is written !i-g, not tU. That is, the root is
written syllabically, and not logographically. The spelling nam-!i-g also occurs in the

Gudea inscriptions. Such spellings show us that the final 11/ of the root was indeed
pronounced in word-final position.

J
Lines 4ff. are a complicated genitive:
nam.til. Ulgi .ak.e
nitabkalaga
lugal. Urim.ak
9. Because the locative-terminative is used in line 3 in the sense ofthe dative, it is resumed
by na, the dimensional-prefix properly belonging to the dative. In its more usual uses, the
locative-terminative is either not resumed at all, or else it is resumed by ni - properly the
dimensional-prefix belonging to the locative.
There is no expressed direct object (patient); the votive cylinder-seal itself is the direct
object.
The last word in the line is an anticipatory genitive, beginning a new sentence: "of this
cylinder-seal, its name is ... ", that is, "the name of this cylinder-seal is ... ". This use of the
anticipatory genitive is similar to that seen in Text 15: alam-ba ... mu-bi-im, "the name of
this statue is ... ".
Line 9 includes the final word of one sentence, and the beginning word of the
following sentence. It is unusual in Sumerian orthography for one line to contain elements
of two different sentences. However, the space constraints in cylinder-seals occasionally
cause odd placement of signs within a line or case.
1 0- 1 1 . These lines express the actual name of the votive cylinder seal.
10. lugal-m l O is probably a vocative. As in Text 2 1 , the vocative is unmarked.
The use of nam to form abstracts has occurred several times. However, some
abstracts are formed with nig, instead of nam: ni g-Sag5 ' "goodness, favor". The original
meaning of nig is "thing", "something". It is relatively uncommon for nig to form such
abstract nouns. More commonly, it forms concrete nouns from verbal roots: g!!7 ' "to eat";
nig-g!!7 ' "food". Therefore, it might perhaps be best to regard the nig-Sag5 in Text 22 as
concrete, rather than as abstract. Sollberger, in fact, translates this expression (in another
text) as: "grace, favour (in a concrete sense, 'good things')" (1966: 1 58).
Unlike nam, nig is not used with nominal roots. That is, such forms as *nig-l!!gal do
not occur.
The transliteration ne assumes the interpretation as .ani.e, the possessive-suffix
followed by the locative-terminative case-marker, here with its more original meaning of
"at, by, through". A similar case occurred in Text 20, where the writing ne represented the
252 Lesson 22
possessive-suffix followed by the ergative case-marker .e. As discussed there, some
Sumero10gists would prefer to transliterate such phrases by ni: getug3 -nig-ag 5-g-ka-ni.
1 1 . g is the modal-prefix of the cohortative mood. This is used for positive wish for first
person, both singular and plural: "let me/us, may I1we". It is thus the first-person
equivalent of the third-person desiderative modal-prefix b.
The view of Sumerian grammar presented in this book assumes the presence here of a
conjugation-prefix 1. which has assimilated into the modal-prefix g. However, since a
writing such as *g-l seems never to occur, the assumption of such assimilation may be
simplistic. Therefore, the idea that a conjugation-prefix must be present in every finite
verbal form may need modifying.
One of the thornier problems in Sumerian morphology is the form of the root
appearing after the different modal-prefixes. In the singular, the cohortative g is always
used with the bamn!-root, whether the root is being used transitively or intransitively. (The
plural is unsure.) When the cohortative is used in the singular, there is no overt marking
for person; that is, the subject (or agent) is not marked. But the direct object (patient),
however, is expressed by a personal-affix immediately preceding the verbal root. Thus, the
.n before the verbal root here marks the patient, and hence the root ti-il must be interpreted
as transitive: "let me keep him alive/well".
The syntax of the personal-affixes of the cohortative (and of other moods) differs from
that of the indicative. In the indicative, the pre-verba1-root slot marks the agent in the
bamn!, but the patient in the mam. But in the cohortative, the agent is unmarked; the pre
verbal-root slot marks the patient, and the root is always in the bamty.
1 2. The enclitic copula is not used, although it was so used in the parallel expression in
Text 9: a1am-ba ... mu-bi-im.
Discussion: sign formation
In Lesson 9, the use of gunu-strokes to form new signs was discussed. A similar
device was the addition of eig-strokes to a sign. Whereas gunu-strokes consist of short
lines, eig-strokes look more like a kind of cross-hatching. For example, the da-sign is
originally a picture of a hand-upper-shou1der-arm: ,i=1. The word has such meanings as
"forearm, side", etc. To indicate .!i, meaning specifically "arm" (and then, by extension,
"strength"), eig-strokes appear on the part of the da-sign which approximately repre
sents the arm, yielding: .
The etymology of eig is unknown, but it must be connected with the word e, "bar
ley", whose cuneiform sign resembles these cross-hatchings.
- Syllabic writings
Syllabic writings of words usually written 10gographically (such as nam-ti-il for nam
til) are not common in the Ur III royal inscriptions; it is not sure why they occur. By way
of contrast, they are not uncommon in the two very large Gudea inscriptions.
Falkenstein believes that the relatively high frequency of such spellings in these two
Gudea texts shows that they were written down on the basis of a dictated text. That is, the
scribe(s) did not work from a written, already-prepared source, but rather the scribe(s)
 Lesson 22 253

listened to the text being read, and wrote down the text as they went along. This idea does
seem like the best explanation for certain kinds of error which occur in the Gudea texts.
There is also some other evidence to support this view.
-Textual interpretation
The interpretation given above of the verb form in line 1 1 rests on the assumption that
the .n in the prefix chain refers to the patient. This view would probably be accepted by
most Sumerologists. However, other Sumerologists are less categorical in their thinking,
and would say that at our present state of knowledge, other possibilities cannot be
excluded. Therefore, this particular line has been translated as: "let me live by his ear of
favor", or even "let me make well his ear of favor". These translations reflect different
possibilities of understanding the .n: marker of the first-person intransitive verb, or marker
of the first-person transitive verb, or marker of the third-person patient, or reduced form of
the dimensional-prefix cross-referencing the locative.
- g and gae
The cohortative modal-prefix is /ga/, and the first person independent pronoun is
/gae/. Since one begins with /g/ and the other with /g/, it does not seem that they are
directly related to each other. On the other hand, the similarity in form, and the fact that
both are used for functions involving the first person, makes one pause. It is less easy,
however, to see such surface-level etymologies with the other modal-prefixes.
- Noun formation
The term zabar-dab5 0riginally meant something like "one who holds the bronze". The
word for "scribe", dub-sar, originally meant "one who writes a tablet". In Text 22, g!!-za
hi, "chair-bearer", occurs. A number of names of officials in Sumerian are composed of an
active participle and an incorporated direct object. Many were borrowed into Akkadian as
simple nouns: zabardabbu, tuparru, guzalU, etc.
- Seals
It is usually assumed that Kilula was the person who had this cylinder-seal fashioned.
He dedicated it to Meslamtaea, to bring life to Shulgi. Presumably Kilula gave it to Shulgi,
who then gave it to the temple. Thus, the cylinder-seal was designed to curry favor with
the king.
It is hard to say why the particular god Meslamtaea was invoked on this seal. Hallo
points out that "in private ex-votos inscribed on behalf of the king, it is not always certain
whether the deity involved is the personal god of the king or his donor" (1966: 1 37 n.53).
Without more precise knowledge of how such dealings took place, it is difficult to say
exactly what the name of the seal does mean. l!!gal, to illustrate part of the problem, can
mean "king, lord", referring to the god Meslamtaea. It can mean "king", referring to
Shulgi. It can even mean "owner", referring to the owner (which one?) of the cylinder-seal.
And, the locative-terminative (presumed ! ) at the end of the nominal phrase can have several
different values: "at, by, through", etc. A somewhat materialistic interpretation of the name
254 Lesson 22
would be: "let me make him well at his ear of favor", that is, "at his ear which hears and
grants favors". The sense is, "let him listen favorably to me", "make him accessible to me".
However, one can't help wondering whether Shulgi had an ear-ache.
- Votive seals
Votive-seals were basically non-functional, in the sense that they were not primarily
designed to be actually impressed upon written documents, as were the functional seals.
Rather, they were votive objects, in the form of a seal. There are very few examples of
votive seals actually being used (although it is always possible that this is an accident of
discovery).
Gelb says:
The main characteristic of the votive seals is that while they identified the
donor of the seal, they were not used by the donor but by the divinity to
whom they were offered. Certain seals can be used for purely ornamental
purposes but nothing would prevent the temple from employing them in
identifying and legal purposes (1977 : 1 12).
Votive seals are also generally larger than functional seals. And, whereas functional
seals are inscribed in reverse - so that the impression comes out correctly - votive seals are
not; they are meant to be looked at, not to be used.
 Lesson 22 255

Text 22a
supplementary
256 Lesson 22
Sign-list and vocabulary

Ija-a-ba-me-er PN. The etymology is unsure, but it is probably Akkadian. Landsber

ger has interpreted it as /baib-amir/, "he was wanted (and then) was seen", two verbal ad
jectives in the predicative state.
U-ku-un-dZuen The name is Akkadian, "Sin has placed/established", Bkun Sin. Its lo
cation is unknown; it may have been close to Ishchali.
Commentary
6. The ki-sign is not in its expected place. Presumably, the scribe wanted to keep the en
sign and the zu-sign together. This particular writing may be regarded as a reflection of the
practice of earlier periods in Sumerian, when the order of signs within a line or case was
not as fixed nor as linear as in later times.
Discussion
According to Hallo, "This is presumably the first Ur III text ever published. It was
first copied by Charles Bellino in 1 820".
 Lesson 23
This lesson reviews and summarizes certain main points of Sumerian grammar, so that
its broad structure will not have become obscured by all the details presented in the
previous lessons. Secondly, specific areas of disagreement among Sumerologists will be
pointed out.
A. General structure
1 . Word-order
Sumerian is basically S-O-V in word order:
NPl -case marker NP2-case marker NP3-case marker VP
(1) (2) (3) (4)
The NP marking the agent usually comes first. The constituents after the agentive NP
are more free; adverbial cases may precede the patient, or the patient may precede the
adverbial cases. Deviations from this standard syntax (for example, a benefactive at the
beginning of a sentence) are for emphasis.
2. NP structure
An NP can be of any length, and can be composed of many different structures: noun;
noun-noun compound; noun+adjective; genitive phrase; noun+possessive-suffix; etc.
These can be combined in different ways. A typical example of a complicated nominal
phrase is: e-gal-zu-!!-ne, for e.gal.zu.ene, "your older brothers".
3. VP structure
The basic structure (omitting some rare optional elements) is:
modal-prefix conjugation-prefix dimensional-prefixes personal-affix
(1) (2) (3) (4)
ROOT personal-affix
(5) (6)

4. Ergativity
It is difficult to see all the manifestations of ergativity in the Ur III royal inscriptions,
because very few intransitive verb forms occur, especially in the maru. Here ergativity will
be summarized, even if some of the constructions do not occur in this corpus.
The following four sentences will illustrate: (1) the bamm-transitive; (2) the maru
transitive; (3) the bam!!-intransitive/passive; (4) the marfi-intransi tivel passive.
bam!!-transitive:
(1) The king built the house.
luga1.e e.0 mu.n.du.0

257
258 Lesson 23

manl-transitive:
(2) The king will build the house.
lugal.e e.0 Lb.du.e.0
bam!!-intransitive/passive:
(3) The king went.
lugal. 0 mu.gin.0
manl-intransitive/passive:
(4) The king will go.
luga1.0 Ldu.0

In both the bamm and the manl, in both the transitive and the intransitive/passive, the
nominal participants are marked in the same manner: the agent is marked in .e (the ergative
case-marker) and the patient is marked in .0 (the absolute case-marker).
Thus, looking at the manner in which the nominal participants in the sentence are
marked, Sumerian is an ergative language: the agent is marked in one manner (.e) , and the
patient in another (.0).
In terms of cross-referencing by the personal-affixes, however, the situation is
different. With a transitive verb, in the bamm the agent is cross-referenced by the personal
affix slot before the root. In the maru the agent is cross-referenced by the personal-affix
slot after the root. In the bam.!! the patient is cross-referenced by the personal-affix slot
after the root. In the maru the patient is cross-referenced by the personal-affix slot before
the root.
With an intransitive/passive verb, the bamm and the maru behave the same way: the
patient is cross-referenced by the personal-affix slot after the root.
This means that the cross-referencing system used in the maru behaves differently in
the transitive than it does in the intransitive/passive. Repeating the previous sentences in a
differen t order,

(5) The king built the house.

lugal.e e.0 mu.n.du.0
(6) The king went.
luga1.0 mu.gin.0
(7) The king will build the house.
lugal.e e.0 Lb.du.e.0
(8) The king will go.
lugal.0 Ldu.0

In the bamt.!! , the patients in (5) and (6) are cross-referenced in the same manner: by
the personal-affix .0 after the root. But in the maru, the patient is cross-referenced in two
 Lesson 23 259

different ways: In (7), the patient is cross-referenced by the personal-affix slot before the
root. In (8), the patient is cross-referenced by the personal-affix slot after the root.
Thus, the patient in (7) (the direct object of the transitive verb) and the patient in (8)
(the subject of the intransitive/passive verb) are not treated the same. Therefore, the system
cannot be considered ergative, since the definition of ergativity is that these two patients be
treated the same.
This argument hinges on the definition of ergativity. In the bamn! and the maru, the
nominal participants are marked in an ergative manner. It is only in the system of cross
referencing that the two patients in (7) and (8) are treated differently from each other. If
one accepts the idea that ergativity does not just refer to the nominal markers, then we are
forced to say that the maru does not behave in an ergative manner. But since the bamtg
does behave in an ergative manner, Sumerian must be called a split ergative language, split
along an aspectual axis. However, if we look only at the nominal markers and not at the
system of cross-referencing, we can say that Sumerian is ergative, and not have to refine
this term (at least, in terms of an aspectual axis; there is some indication that Sumerian is
split along a nominal - pronominal axis).
(5) Personal-affixes
The interpretation of the personal-affixes presented here has been basically known
since Poebel. This interpretation was made much more explicit, and placed in an ergative
framework, by Michalowski (1982). While this interpretation seems to work for most
verbs, it is clear that it does not work for all of them. For example, a commonly occurring
sentence is: kiib PN.ak ib-ra, "the seal of PN was rolled". The verb apparently is intran
sitive/passive, as shown by the lack of an ergative case-marker .e, yet the .b in the pre-root
slot seems to cross-reference it. The proper analysis of such forms is still unclear to us; it
may be much more complicated than it first appears. More work remains to be done on
explaining and categorizing the exceptions which seem to occur with the personal-affixes.
(6) Root
Sumerian roots can be divided into two classes: nominal roots, such as lu, and verbal
roots, such as sar; there is no morphologically distinct class of adjectival roots. Adjectives
are to be regarded as participles (or something similar) of verbal roots. For example, gibil
can be used as an adjective meaning "new", from the verbal root meaning "to be new".
There is no canonical shape of the verbal root. Most of the verbal roots occurring in
the texts in this book are of the type CV (d.!!, g4), or of the type eve (g!!l2, pad). How
ever, there has been one eveve root (Wig), one of the type vev CM!!), one unsure (ba
ill) , and one borrowed from Akkadian (gi-in). Similarly, there is a fairly wide variation in
the shape of the nominal root. They have taken the form V @; CV Clli!.) ; VC (ur); eve
(gg); vev (utu); evev (dumu); veve (alam); eveve (temen); etc.
The root is unmarked for such categories as active - passive, transitive - intransitive,
causative - passive; etc. For example, til can mean "to live" or "to let live"; kur9 can mean
"to enter" or "to make enter", "to bring in". In Text 15, gub is used in the meaning "to
stand"; in Text 6, it is used in the sense of "to plant" a garden.
260 Lesson 23

B. Areas of disagreement
There are a number of disagreements and alternate explanations about certain features
of Sumerian grammar. Because a knowledge of these alternate views is presupposed by
Sumerologists, it is important to be at least familiar with the main differences from the
views presented here.
1 . Verbal phrase
The interpretation of the Sumerian verbal phrase presented in this book owes much to
the ideas of Gene Gragg (briefly sketched in [1968]). Two principal differing views are
those of Falkenstein and Jacobsen. Yoshikawa's views also differ considerably, but he has
not yet published a complete synthesis of his views.
(a) Falkenstein
His views are sketched in (1959), and particularly adumbrated for the Gudea texts in
(1978 2). His interpretation of the prefix chain is as follows:
Prllformative Konjugationsprllfixe Prafixe Verbalinfixe Wurzel
(1) (2) (3) (4) (5)

Pr!lformative: Correspond to the modal-prefixes.

Konjugationsprafixe: Include only the following: l.. mu, and al.
Prafixe: Correspond to the dimensional-prefix na, and the conjugation-prefixes ba and
bi.
The conjugation-prefixes limmi/, limma/, etc., are considered by Falkenstein to
represent *i-bi-i and *i-bi-a respectively; that is, they contain both a Konjugationsprafix and
a Prafix.
Verbalinfixe: Include the dimensional-prefixes except na, and the personal-affixes.
Falkenstein's views of Sumerian grammar have been criticized on two basic grounds:
a mixing of synchronic and diachronic description, especially in the categories of Kon
jugationsprafixe and Prafixe, and a regular use of terms and concepts proper to the Indo
European and Semitic languages, but which do not necessarily apply to Sumerian, or to
languages which are typologically similar to it.
(b) Jacobsen
His views are spelled out most explicitly in (1965); some important comments occur in
his review of Thomsen ( 1988). His terminology for the prefix chain is based on formal
criteria:
Profixes Prefixes Infixes Root
(1) (2) (3) (4)
 Lesson 23 261

Profixes: Correspond to the modal-prefixes.

Prefixes: Correspond to the conjugation-prefixes. However, Jacobsen has definite
and specific views on the meaning and function of the conj ugation-prefixes. He believes
that they can be analyzed into two morphemes: "a pronominal-adverbial element denoting a
region (m, b) and a relater or case-mark (u for tangentive, a for illative, i/a for allative)."
This analysis is very detailed.
Infixes: Correspond to the dimensional-prefixes and personal-affixes, in a rather
complicated way.

In general, Jacobsen's analysis tends to be much finer than that of other Sumerologists.
In some cases, vocalic (or consonantal) alternations which most Sumerologists would
regard as phonetic or morphophonemic alternation are regarded by Jacobsen as repre
senting different morphemes. He has been criticized for producing much too detailed a
segmentation of the morphology. On the other hand, Jacobsen's interpretations are based
on a close empathy for the texts; perhaps more so than any other living Sumerologist,
Jacobsen has a feeling for and understanding of the content of Sumerian texts, especially
literary texts.

(c) Yoshikawa

Yoshikawa has published a number of articles in an attempt to define the functions of

the conjugation-prefixes. In some of the later articles, he has tried to use phenomena from
other ergative languages, especially Georgian, to elucidate Sumerian. In his 1981 article, he
says:
We can tentatively specify the function of the respective Sumerian verbal
prefixes as follows:
1. bi -: Ientive locative/superessive prefix
im-mi- : Ventive locative/ superessive prefix
ba- : Ientive reflexive (/subjective) prefix
im-ma- : Ventive reflexive (/subjective) prefix
II. mu-: Topical agentive (/objective) prefix
i - : Non-topical agentive (lobjective) prefix
III. al- : Neutral prefix.

2. Maru-inflection

maru-forms such as l-sar-re have been analyzed here as Lsar.e.0. The root is sar,
which is a member of the affixation-class of maru-formation; .e is the maru-suffix; .0 is
the personal-affix cross-referencing the agent, that is, the third person marker.
Edzard believes that l-sar-re is to be analyzed as Lsar.e. sar is a member of the in
variable class of maru-formation, and .e is the third person marker.
The first analysis sees two morphemes after the root; the second sees one. It is not
easy to resolve this issue. The problem is partially the result of a lack of evidence. Far
fewer intransitive/passive verbal forms are preserved than transitive forms. Far fewer
present-future fornls are preserved than past. And, fewer first and second persons are
262 Lesson 23
preserved than third, and fewer plural forms than singulars.
Many of the examples preserved which show the greatest amount of grammatical
variation are attested in relatively late copies of literary texts. In these texts one must
always guard against misunderstandings by the Akkadian-speaking scribes who copied
down these texts. Similarly, the Akkadian-speaking scribes who drew up the grammatical
texts discussed in Appendix 2 sometimes include transitional or analogical forms, which
cannot be regarded as conforming to the standards of earlier Sumerian.
The problem is also caused by ambiguities in the writing system. Types of ambi
guities include: (1) The writing system cannot distinguish between a form such as i .sar.en
or L sar.e.en; they would both be written as l-sar-re-en. Isolated writings of the type i-sar
re--en do not resolve such ambiguities, because there are several possible explanations for
the full writing. (2) /d! is an amissable consonant. Therefore, a writing such as i-sar-re
can stand for i .sar.e, i . sar.ed, or L sar.e.ed. (3) The de-sign is ambiguous; it can stand for
de or for ne. This means that a writing such as i-sar-re--de can represent several different
possibilities: Lsar.ed.e., L sar.e.ed, Lsar.ene, L sar.e.ene, etc. It is ambiguities such as these
which make it diffic ult to correctly analyze the morphology of the maru.

3. Manl classes

The number of different types of formation of the maru from the bamt!! is unsure.
Yoshikawa's initial formulation showed three classes: (1) affixation (2) reduplication (3)
alternation. This scheme works well for the Ur In royal inscriptions, but this is partially
because only a very limited number of man1 forms occur in these texts. Yoshikawa himself
has said that his classification system will need expansion and revision.
Edzard modified and extended the system into five classes: (1) unchanging. Since
Edzard considers .e to be a third person subject marker, not a marker of the marG, this is
actually the same as Yoshikawa's affixation group. (2) reduplication (3) root-varying.
This is a sub-class of Yoshikawa's alternation class; the two roots are different, but phone
tically similar in some way. (4) alternation (5) irregular. These last do not seem to fit into
the first four classes.
As more progress is made in Sumerology, the last class will be further refined. Some
of the verbs which now seem irregular will eventually be shown to follow rules which are
not yet known. Whenever linguists study an unknown language, they are apt to see more
irregular forms than when they have been able to examine the data more thoroughly.

4. Normalform

There has been some discussion about the conjugation of the intransitive/passive verb
in Sumerian. Poebel believed that there was one conjugation for the bamt!! intransi
tive/passive, and a different conjugation for the marG intransitive/passive. Most modern
Sumerologists, however, believe that there is only one conjugation for both the bam n! and
the manl (that is, there is only one set of endings), although the particular root used (i.e.,
either bamn! or marG) will differ. This one conj ugation is usually called the "Normalform"
(even though this is a rather meaningless term). What follows is Poebel's reconstruction of
the bam n ! intransitive/passive and the marG intransitive/passive.
 Lesson 23 263

For Classical Sumerian, Poebel reconstructs the intransitive/passive form of the bamt.!!
as:

first person singular i-sar-en i . sar.en

second i-sar-en L sar.en
third i-sar i . sar.0
first plural i-sar-re-en-de-en i .sar.enden
second i-sar-re-en-ze-en L sar.enzen
third i-sar-re-eS L sar.eS

He reconstructs the intransitive/passive of the maru as:

first person singular i-sar-re-de-en L sar.ed.en

second i-sar-re-de-en L sar.ed.en
third i-sar-re L sar.ed.0
first plural i-sar-re-de-en-de-en i . sar.ed.enden
second i-sar-re-de-en-ze-en i . sar.ed.enzen
third i-sar-re-de-eS L sar.ed.eS

That is, the man} intransitive/passive differs from the bamm intransitive/passive sole
ly in the presence of the element .ed. Poebel believed, however, that this sytem worked
only for Classical Sumerian. Under the influence of Akkadian, several analogical changes
took place in the system of the man} intransitive/passive, eventually yielding a completely
new paradigm:

first person singular i-sar-re-en i. sar.en

second i-sar-re-en i .sar.en
third i-sar-re L sar.e
first person plural i-sar-re-en-de-en i .sar.enden
second i-sar-re-en-ze-en L sar.enzen
third i-sar-re-ne i .sar.ene

The .ed element has disappeared; the third person singular is now marked by .e; and
the third person plural is now marked by .ene. Thus, the endings have been assimilated to
those of the maru transitive.
From the period between the time of the paradigm of the more classical period and the
time of the newer paradigm, several transitional forms are attested (for example, i-sar-re-de,
L sar.ed.e) for the maru intransitive.
 PART THREE: APPENDICES

Appendix 1
History

The periodization of Sumerian, like the periodization of Akkadian, is to some extent

based on non-linguistic criteria, such as political and historical events. The essence of this
periodization is:
I. Archaic Texts (3 1 00 BC)
IT. Archaic Sumerian (3 100-2600 BC)
ITI. Classical Sumerian (2600-2300 BC)
IV. Neo-Sumerian (2300-2000 BC)
V. Post-Sumerian (2000 BC- l OO AD)
I. Archaic Texts (3 1 00 BC)
The earliest known tablets containing writing come mostly from Uruk (whose modem
name is Warka). They come from a stratum usually designated as Uruk IVa, commonly
dated to about 3 1 00 BC. They were first assigned to "Uruk IVb", but then redated to
"Uruk IVa" (later than Uruk IVb). Secondary literature often confuses this, and so some
accounts speak of Uruk IV, some of Uruk IVa, and some of Uruk IVb.
Almost all of the tablets were found in a "dump" inside the Eanna temple complex.
This means that it is very difficult to date the tablets, either in terms of relative chronology,
or in terms of absolute chronology. According to Eva Strornrnenger, a further complication
is the fact that to some degree, the architectural levels and building phases at Uruk have
been dated on the basis of the tablets found, not vice-versa. (This is also true of some
tablets found later than Uruk IV.) She says that
The phases of writing are everywhere used in order to date the level in which
the specific tablets were found. This leads to the conclusion that the
possibility of the existence of tablets prior to IVa ... has not been recognized,
but that deposits in which a few tablets were found have been dated to IVa
on that basis alone (1980:48 1).
Other scholars, however, are less willing to agree with her implicit criticism of the
reconstruction of the archaeological history of the site.
The first of these texts were excavated by the German archaeologists of Uruk during
1 928- 1931. In 1 936, Falkenstein published his Archaische Texte aus Uruk, a seminal
work. In it he treated the first 620 tablets found. By now, upwards of four thousand of
these texts are known, found during subsequent excavations. The majority of these (mostly
fragmentary) tablets have not yet been published.
There are a few similar tablets from other sites, mostly in northern Syria. In addition,
both from Uruk and especially from Syria (and elsewhere, including Iran) there are a
number of what are commonly called "numerical" tablets. It was at first thought that these

265
266 Appendix 1

tablets represented numbers, but more likely they indicate items being counted; their precise
intepretation is unsure. Two from Tell Brak in northern Syria, found in 1984, in particular
are quite archaic looking. Based on archaeological criteria, these numerical tablets do not
seem to be any older than the pictographic tablets discussed above; rather, both numerical
and pictographic tablets occur in Uruk IV a.
A very archaic-looking tablet comes from Kish. Unfortunately, it is from an
uncontrolled context, and so it cannot be dated archaeologically. This tablet is frequently
referred to in popular literature about the Ancient Near East as being one of the very earliest
tablets known (if not the earliest tablet), but it has a fairly elaborate division into cases,
which makes it more likely that it is later than Uruk IV.
These early texts are undeciphered, and perhaps to some degree undecipherable.
Therefore, it is impossible to be certain about what language they are written in. There are
several reasons why these texts cannot yet be read:
- Even at this early date, the supposedly pictographic nature of the signs is not always
obvious. Most of the signs are already abstract; it cannot be determined what they were
originally meant to depict.
- Some of the abstract signs can be understood on the basis of knowledge of later
Sumerian. However, a fair number of the signs (perhaps 30%-50%) cannot be read or
understood. These are signs which eventually passed out of use, so that there is no later
grammatical tradition to provide information about their meaning.
- Most of the signs which can be understood are logographic; in theory, these can be
read in any language. A sign which is a picture of a mountain, for instance, could be read
as "mountain", "Berg", adfi, kur, etc.
- There do not appear to be any syllabic signs. This means that no grammatical features
can be seen; for example, there are not any case-markings on nouns. There do not appear
to be any verbs at all. Writing at this stage was a highly mnemonic device.
- These are mostly administrative records, sometimes very short - occasionally, just a
few signs long. It is very difficult to understand such texts out of context, that is, without
knowledge of the administrative framework which produced these texts.
In spite of these problems, most scholars think that these texts are written in Sumerian.
The main reason is because texts have been preserved from the later Uruk III stratum,
which is known to be Sumerian. Since archaeologists see a cultural continuum between
Uruk IV and Uruk Ill, it is reasonable to assume that the same language is present in both
strata. In addition, Powell has argued that the system of metrology used in these early texts
seems to be the same system used in clearly Sumerian texts:
The system of numeration deducible from the notation present on Uruk
IVa/III tablets makes it virtually certain that these tablets are written in
Sumerian and, ipso facto, highly probable that the inventor of the pictorial
writing system was also a Sumerian (1981 :423).
These tablets are being studied by Margaret Green and Hans Nissen (a student of
Falkenstein), and are in the course of publication (preliminary discussions Nissen 1985,
1986; first major publication Green and Nissen 1987). Although most earlier scholars
differentiate between Uruk IVa and Uruk Ill, Nissen subsumes both into one category,
 History 267

"Archaic Texts". (The figure of "4000 texts" cited above therefore includes tablets from
both Uruk IVa and Uruk lll.) Nissen estimates that about 85% of the texts are economic
records, and about 15% are lexical lists. He is more optimistic than most scholars about the
possible decipherment of these texts, believing that he can identify about 700 of 1000
different signs, and that the texts are "possibly" written in Sumerian. He bases his
arguments partially on the continuity between the early lexical lists (unknown at the time of
Falkenstein's publication) and later, well-understood lexical texts.
Nissen thinks that the texts from Uruk can be divided roughly into two classes, one
representing an "early stage of the script", and one a "younger stage of the script" (these
two divisions do not exactly correspond to the traditional Uruk IVa/Uruk III divisions).
The latter, consisting of most of the tablets found since Falkenstein's publication, are more
amenable to analysis.
The fact that many of the signs in these early tablets are already abstract has led many
scholars to assume that there was some previous development behind the signs. That is,
these tablets do not represent mankind's first attempt at writing. Several different
hypotheses have been proposed:
- The Sumerians may have borrowed their writing system from some other people,
perhaps some distance away from Mesopotamia. This is not impossible. It has often been
argued that the Sumerian writing system does not fit the Sumerian phonological system
very well; this might imply that the writing system was created for a different language.
This particular theory has been around for many years; it is obviously very difficult to
prove.
- Earlier writing may have been on perishable material, such as wood, or animal skins,
or palm leaves, etc. There are parallels to such practice from later Mesopotamia, and from
Arabia around the time of Muhammad. This is also a rather old theory, but it is also
virtually impossible to prove.
- In a series of articles beginning in 1977, Denise Schmandt-Besserat has argued that
the earliest "precursor" of writing was clay "tokens", which have been found at various
sites throughout the Ancient Near East, starting from the early Neolithic. Writing
originated in a conceptual leap, from the use of physical tokens, to the use of symbols to
represent these tokens: "The substitution of signs for tokens was no less than the invention
of writing" ( 1986:37). She envisages the following stages (1986:35):
(1) 8000 BC appearance oftokens
(2) 3250 BC clay envelopes hold tokens of particular transactions
(3) 3200 BC signs are impressed on the surface of envelopes
(4) 3 100 BC clay tablets appear with impressed and incised signs
Powell agrees:
Cuneiform was invented in a short period of time around 3000BC by a
citizen of the Sumerian city of Uruk. ... It arises conceptually out of the token
system described by D. Schmandt-Besserat. ... The pictorial ancestor to
cuneiform writing was invented as a conceptual whole during the time period
represented by the Uruk IV-Ill archaeological strata (198 1 :419-420).
268 Appendix 1

Contrast this with the more "evolutionary" thinking of Walker:

Thus it is beginning to look as if we should think in terms of the invention
of writing as being a gradual process, accomplished over a wide area, rather
than the product of a single Sumerian genius ( 1986:9).
Nissen hints at the complexity of the developing of writing: "Writing was developed
at the end of the fourth millennium B.c. by a mixed language group in which Sumerian
was apparently the main component" (1988: 14).
The most vocal critic of Schmandt-Besserat's view has been Lieberman (1978, 1980).
However, probably the majority of Sumerologists agree with her overall interpretation of
the development of writing from tokens.
11. Archaic Sumerian (3100-2600 BC)
Tablets from this stage are found at several sites. The oldest are from Jemdet Nasr;
others are from Uruk Ill, Uruk 11, Ur, Fara (ancient Shuruppak) , and Tell Abu Salabikh
(2600 BC). These dates are not exact, and, as discussed above, Nissen includes both Uruk
IVaJUruk III together; the ramifications of this revised relative chronology are still to be
worked out.
In the Jemdet Nasr texts, there is a personal name written: dEn-lil-1!. The word for
"arrow" in Sumerian was ftil; the sign used to represent this word was originally a picture
of an arrow. It is doubtful, however, if this name means something like "Enlil is an arrow".
But, the root ftil in Sumerian also means "to live". Thus, the name dEn-lil-1! means "Enlil
lives", or more likely "May Enlil give life", "May Enlil keep alive". That is, in this name a
cuneiform sign is being used syllabically: The 1i-sign is being used strictly for its phonetic
value, not for its logographic value. This writing shows that the script is being used for a
language where the words for "arrow" and "to live" are homophonous.
This interpretation of the personal name dEn-lil-1i goes back to Falkenstein, in 1936.
It gained immediate acceptance by Assyriologists and Sumerologists, who believe that the
writing: (1) shows the existence of phoneticization; (2) shows the personal name to be
Sumerian; and (3) shows the language of the texts in which the personal name occurs to be
Sumerian. (The same name may also occur in the later phase of Nissen's "Archaic Texts".)
In 1974, however, A.A. Vaiman suggested that the name should be read as E.EN.TI; he is
followed by Lieberman.
It is possible that there are other instances of phoneticization in the texts from Jemdet
Nasr, but the evidence is not unequivocal. Texts from the later stages of this period show
increasing phoneticization, and are clearly Sumerian.
Up until the 1960s, virtually all of the texts which were known from this period were
the usual administrative and economic texts. These are not always easy to understand,
again because of the lack of any context. This situation was changed in 1963. In that year,
the University of Chicago began excavations at a site called Tell Abu Salabikh, near
Nippur. It turried out that the majority of tablets and fragments found were literary texts.
Some were compositions which were known from later times. For example, there is a text
known as the "Kesh Temple Hymn", preserved in numerous Old Babylonian copies dating
to about 1 800 BC. Fragments turned up at Tell Abu Salabikh - some eight hundred years
 History 269

earlier. Other texts turned out to be previously unknown compositions. For example, one
is a collection of temple hymns. Most of these texts are scarcely intelligible; not much is
known about literary Sumerian of this period.
Thus, the primary importance of Tell Abu Salabikh lies in the existence of literary texts
from the middle of the third millennium B C. Since these discoveries, scholars have
recognized fragments of literary texts among some tablets which have been known for
many years. For instance, some of the Fara texts are fragments of proverbs which are
known from later proverb collections.
The texts from Tell Abu Salabikh are also important, because a number of the literary
texts (and lexical texts) have colophons of the sort: "so-and-so wrote". It is not known
what the word "wrote" means here exactly: Does it mean that the scribe "composed" the
composition, or that the scribe "copied" the text from a master tablet, etc. However, what is
interesting is that a number of these scribes have demonstrably Semitic personal names.
It is difficult to date the intrusion of Semitic-speaking peoples into Mesopotamia, on
linguistic or other grounds. The first evidence is usually thought to be the presence of
Akkadian loan words in early Sumerian. These loan words are difficult to evaluate, how
ever, because it is not always certain which way the borrowing went, or whether a third
language may have mediated a word, etc.
The Semitic names in these colophons are thought to be the first real evidence of
Semitic-speakers in Mesopotamia. If the Fara texts and the Tell Abu S alabikh texts are
dated to about 2600 BC, that gives a terminus ante quem for the arrival of Semitic speakers,
but it does not say anything about how long they might have been present in Mesopotamia.
If they had already become scribes, they must have been there for some time, since they had
worked themselves into the intellectual life of the community.

Ill. Classical (or Old) Sumerian (2600-2300 BC)

Most of the texts of this stage come from Lagash, from a period known as the "First
Dynasty of Lagash". Besides the usual administrative, economic, and legal texts, there are
a fair number of royal and private inscriptions. There are also some letters, and even a few
literary fragments are now known. Royal inscriptions are also known from other sites.
The end of this period corresponds to the rise to power of the Semitic-speaking
Dynasty of Akkad (2334-2 1 54 BC). As mentioned above, Semitic-speaking peoples must
have been present in Mesopotamia for centuries before the time of Sargon, the founder of
the dynasty (ruled 2334-227 8 BC). It must be presumed that Mesopotamia was bilingual
during this time, at least to some degree. However, with a Semitic-speaking dynasty in
power, Sumerian gradually started to move into second place.
A recent addition to the corpus of texts known from this period are the texts from Ebla
(in northern S yria, therefore from outside of the Sumerian-speaking heartland). To date,
upwards of ten thousand texts and fragments have been discovered. The texts are in both
S umerian and Eblaite; until the material is better studied, it is not sure which language
predominates. Early accounts said that perhaps ninety percent of the texts were written in
Sumerian; this is probably much too high a figure. The problem is that the texts written in
Eblaite are couched in a Sumerian orthography, utilizing a large number of Sumerian
270 Appendix 1

logograms.
Most of the texts found at Ebla are administrative or economic, chiefly concerned with
the metal and textile industries. However, there are also lexical lists; some are Sumerian
lists known from later periods, others are bilingual Eblaite-Sumerian texts. There are a few
literary texts in Eblaite (mostly incantations); these are extremely difficult to understand.
The existence of possible literary fragments in Sumerian is disputed.
It is still too early to assess the Sumerian texts from Ebla. It is clear, however, that
much new information is present. For example, the bilingual lexical texts include Sumerian
words and expressions not elsewhere attested.

IV. Neo (or New) Sumerian (2300-2000 BC)

Although Sumerian was on the defensive in the face of Akkadian, it enjoyed a strong
- albeit brief - revival under the kings of the Vr ill Dynasty (2 1 1 2-2004 BC). This is the
period from which the most tablets of all have been preserved. There are texts from many
sites, including Vr itself, Drehem, Lagash, Larsa, Nippur, and Vmma. There are literally
thousands and thousands of mostly economic documents, as well as inscriptions, letters,
and other types of texts. Also, more and more tablets with literary texts are being dated to
this period.
From some time before the reign of Vr-Nammu (the founder of the Dynasty), there are
a fair number of inscriptions from the reign of Gudea, the local ruler of Lagash. Many of
these are inscribed on statues of Gudea himself. There are also two large cylinders of his,
inscribed with a very long building hymn. The largest ("Cylinder A") is almost one
thousand lines long; it is apparently the longest connected Sumerian inscription.
The dynasty of Gudea is referred to as "Lagash II". The chronology of Lagash II is
unsure; some see it as roughly contemporaneous with Vr Ill, but most view it as following
immediately upon the Old Akkadian period. In any case, the language of the Gudea texts is
more or less the same as that of the Vr III texts.
It is not known when Sumerian ceased to be a spoken language; this is a current topic
of discussion among Sumerologists and Assyriologists. Vsually assumed to be spoken
during the Vr III period, it was under the greatly increasing influence of Akkadian. Some
scholars use the figure 2000 BC, others 1 900 BC, for the date when Sumerian ceased to be
spoken, but this figure is rather arbitrary. The language continued to be spoken by ever
smaller groups of speakers, and it is impossible to say when the last speaker of Sumerian
died. What is usually meant by the question "When did S umerian die out?" is "When did
the native language of the people who produced the texts we have cease to be Sumerian?"
Pockets of native speakers of Sumerian may have continued for some time, but without
producing any texts.
Other scholars have argued for an earlier death. Jerrold Cooper has said that
"S umerian as a spoken language was in all probability dead or nearly so in Ur Ill"
( 1 97 3 : 24 1 ). His argument is based on the types of documents preserved during the Ur III
period. Both Kienast ( 1 98 1 a) and Michalowski ( 1987) essentially agree.
Jacobsen, on the other hand, says "We therefore assume that Sumerian was still
spoken as everyday language in the south in the Ur ill period and a major part of the Isin-
 History 271

Larsa period as well ( 1988: 124). Liebennan has stated that there is some evidence to show
that "Sumerian was spoken during the Old-Babylonian period" (1979:27).
V. Post-Sumerian (2000 BC-lOO AD)
This is occasionally divided into the following subdivisions:
Early Old Babylonian 2000- 1800 BC
Later Old Babylonian 1 800- 1600 BC
Post Old Babylonian 1600-
After the Early Old Babylonian period, Sumerian was essentially dead as a living
language. However, it continued to be taught in the schools as a language of culture, and as
a language of religious importance. The parallel has frequently been made with the role of
Latin in the Roman Catholic Church: Latin is still written, and even to some degree
spoken; hence, it is "living", even though it is not spoken as a native language of anyone.
Recently, Vanstiphout has used the tenn "Standard Sumerian" to mean
the language used in the literary documents of the Ur III and Old Babylonian
periods. ... This language is a literary and therefore written fonn, taught in
school for educational and literary purposes ( 1985: 1 ).
The majority of Sumerian literary tablets which have been preserved are from the Early
Old Babylonian period. However, there are also original texts written in Sumerian from
this period; examples include royal inscriptions (alongside those written in Akkadian), and
hymns written in honor of some of the Old Babylonian rulers.
Although the original composition of most Sumerian literary texts was in Sumerian,
by native speakers of Sumerian, the native language of the scribes who copied down the
literary texts during this period was Akkadian, not Sumerian. This led to a strong linguistic
influence of Akkadian upon Sumerian, to the extent that the literary texts contain features
which would appear to be "wrong" by the rules of Classical Sumerian grammar.
Michalowski, for example, has spoken of the "profound changes in grammar evident in the
Old Babylonian literary texts" (1980a:9 1); "during the Old Babylonian period
Mesopotamian scribes wrote Sumerian utilizing a profoundly different grammar, much
influenced by Akkadian" (1980a:86 n.3). The extent of deviation from the nonn varies
from one particular text to another. Inanna's Descent, for example, is pretty good Su
merian, with only a few "wrong" verbal fonns. In Gilgamesh and Agg, on the other hand,
there are more "wrong" verbal fonns than "right" ones.
It is, of course, always possible that fonns which we regard as "wrong" are in fact
"right", but our understanding of Sumerian grammar is not yet sophisticated enough to
correctly interpret such fonns. J acobsen has emphasized this methodological point:
Once it has been decided that our sources are generally suspect it becomes
natural to see all unexpected and difficult features as due to corruption,
without seriously considering the possibility that our own limited and rough
knowledge might be at fault and need nwision . ... The essential thing is to be
slow to dismiss difficulties with the easy assumption of mistakes by the
Ancients ( 1988: 1 25-1 26).
272 Appendix 1

Sumerian continued to be written right down to the Christian era. These very late texts
are either cultic or astronomical. There are even a few Sumerian texts (including portions
of canonical lexical lists) written in Greek characters. The very latest cuneiform texts
preserved are several astronomical almanacs, written in Akkadian with a great number of
Sumerian logograms for technical terms. The latest of these can be dated by internal criteria
to the year 385 of the Seleucid Era, corresponding to 74175 AD.
As mentioned above, this periodization (and most others) is to some degree based on
external (historical and political) criteria, not on purely linguistic criteria. Jacobsen has
proposed a different scheme, based on linguistic criteria (without yet assigning precise
dates), while emphasizing the fact that the paucity of the data prevents overly-fine
subdivisions:
I Archaic
IT OldSumerian
ITI Standard Sumerian (beginning with Naril.m-Sin of Akkad)
IV LateSumerian
 Appendix 2
Mesopotamian Sources

Much of our knowledge of Sumerian derives from the intellectual activity of the
Mesopotamian scribes themselves. This section describes some of these Mesopotamian
sources.

Lexical lists
It is especially in the area of lexicography that modem Sumerological studies depend
on native sources. From a very early period, the Sumerians began to compile "lexical lists".
These early texts were monolingual, consisting simply of lists, usually of words for
semantically related things: lists of names of fishes, of professions, of stones, etc.
Although most lexical lists are loosely arranged according to subject, others are organized
according to graphic shape, or even according to phonological shape. Fragments of such
lists occur among the very earliest Sumerian texts which have been preserved.
These early texts were the product of Sumerian scholars, originating in the Sumerian
scribal school system. Lexical lists become more and more common, however, beginning
about with the Old Babylonian period. At that time, Sumerian was in the process of
completely dying out as a spoken language - if it had not already done so. By the end of
the Old Babylonian period, if not earlier, Sumerian was only spoken in the schools. These
later lexical lists are a product of the Mesopotamian scribal schools; their purpose was to
aid the Akkadian-speaking scribes in their study of Sumerian.
By the late Old Babylonian period, many lexical lists assumed what is often called
"canonical" status; that is, they became standardized in content and in form. There are
about a dozen such "canonical series". Some are monolingual in Sumerian, like the earlier
texts, but most are bilingual; they have a Sumerian word in the left-hand column, and an
Akkadian equivalent in the right-hand column. Some have three columns: a phonetic
spelling of the sign; the S umerian logogram; and the Akkadian meaning.
Many of these series are quite extensive. One of the largest and best-preserved is
known (both to us and to the ancient Mesopotamian scribes) as "urS -ra = bubullu", after its
first entry. In its canonical form, this series occupied 24 large tablets, totalling about
1 0,000 entries. Civil has called it an "inventory of material culture" ( 1 976: 1 25). He
describes its contents as:

(tablets 1 and 2): legal and administrative terminology.

(3-7): trees and wooden artifacts.
(8-9): reeds and reed artifacts.
( 1 0) : pottery.
( 1 1 ) : hides and copper.
( 1 2) : other metals.
( 1 3) : domestic animals.
( 1 4) : wild animals.

273
274 Appendix 2

(15) : parts of the body.

( 1 6): stones.
(17): plants.
( 1 8): birds and fish.
( 1 9) : textiles.
(20-22): geographic tenns.
(23-24) : food and drinks.

The first entry of this series has Ufs-ra (the Sumerian word for "interest-bearing loan")
in the left-hand column, and the Akkadian gubullu (with the same meaning) in the right
hand column.
Lexical lists such as these help us to detennine the meaning of Sumerian words. Some
of the lexical lists go even further, and enable us to determine the reading (that is, the
approximate phonetic rendering) of a certain sign. For example, there is a relatively late
lexical series known as "diri", which in its canonical fonn occupied seven tablets, with
more than 2,000 entries. This series was used to give the pronunciation of compound
logograms, that is, logograms whose reading cannot be inferred from the individual parts
(such as zabar, written with the UD-KA-BAR signs; without lexical lists, it would be vir
tually impossible to deduce that the pronunciation of these three signs was Izabar/). For
this reason, such compound logograms are often referred to as "diri-compounds".
In diri, the pronunciation of the logogram under discussion is given in the far left-hand
column, using a restricted number of syllabic signs. Then comes the logogram in question.
Then comes the name of the sign (at least as early as the Old Babylonian period, the
Akkadian scribes gave names to the individual signs). Finally, the last column gives the
meaning of the sign, in Akkadian. A typical entry reads:
di-ri diri si-y-ku wa-at-ru
This tells us that the sign is read Idi-ri/. Graphically, this sign "looks like" the si
sign ( ) followed by the -sign ( Tf ). (At least, in this period of cuneifonn writing.
Originally, the diri-sign may have had no connection at all with either the si-sign or the
sign. However, by the Old Babylonian period, when signs were becoming more linear, it
happened to assume a shape looking like the si-sign followed by the -sign.) Because of
this external similarity, the Akkadians named this sign "the si of ", that is, si.a. (k). Finally,
the last column gives the Akkadian translation, "excess" or "extra".
Copies of these canonical texts have been found all over the Near East, not just in
Mesopotamia. There are also somewhat similar texts, but not of any canonical status, both
from Mesopotamia and from outside of Mesopotamia. Their function was the same, to aid
local scribes in their mastery of S umerian (or of some other language).
Some of these non-canonical texts are bilingual, some are trilingual, and some even
quadrilingual. For example, from Boghazkoy in Asia Minor there are several Sumerian
Akkadian-Hittite trilingual vocabularies. The native language in Boghazkoy was Hittite;
these texts were designed to help Hittite scribes in learning both Sumerian and Akkadian.
From Ugarit, there is a quadrilingual "vocabulary". It has entries in Sumerian, Akkadian,
 Sources 275

Hurrian, and then Ugaritic (written syllabically).

The finds at Ebla have produced a new, important source of lexical texts of different
kinds. Some are related to the later Mesopotamian tradition; some are independent
creations. Many are monolingual in Sumerian (less frequently in Eblaite). Others are
bilingual, with the Sumerian again on the left and the Eblaite on the right. In some cases,
the pronunciation of the Sumerian is given, using a reduced number of syllabic signs.
Lexical lists occur among the oldest tablets known. Nissen estimated that some 15%
of the Archaic Texts from Uruk are lexical texts. One, the "Standard Professions List", is
well-known from later copies; the Uruk version lists some 100 different professions and
titles.
Lexical lists continued to be used up to the very last stages of cuneiform. Copies of
some of the canonical lexical series are known from as late as Seleucid times, when
Akkadian itself was no longer a spoken language, having been replaced by Aramaic.
The Mesopotamian lexical lists are not always easy to use. Alongside errors of va
rious kinds which have crept in - some due to the normal accidents of textual transmission,
some due to Akkadian scribes not understanding their originals - there are several con
ventions and abbreviations used by the scribes, which make it difficult to understand the
texts. Also, these lists should be thought of as a kind of "bare bones" text; there was
undoubtedly a tremendous amount of oral information passed on in the Mesopotamia
schools, fleshing out these texts. Unfortunately, there are only hints of such oral teaching.
In addition, it is not always easy to recognize the order of entries in the lexical texts.
As Civil has said,
When they attempted to make an inventory of Sumerian words, the native
Mesopotamian scribes faced a problem familiar to any lexicographer in the
first stages of planning a dictionary: should the entries be organized
thematically, by subjects, or should they be arranged in a serial order based
on graphic or phonological characteristics of the words? One can hardly
speak of planning in the compilation of the Mesopotamian lexical lists as a
whole, since they were the result of a slow process, which lasted for
centuries and answered many different kinds of needs: scribal training,
interpretation of traditional texts, composition of new texts, and, un
doubtedly, a certain amount of simple philological curiosity, spurred on by
the desire of salvaging the words of an extinct language. Nevertheless, the
compilers of each new addition to the traditional lexicographic corpus had to
decide how the entries should be arranged (MSL XIII [197 1] 3).
Civil (1976) has written a most useful article listing and discussing the most important
Sumerian lexical texts. He also discusses the modern editions of these texts, and the
theoretical principles which need to be observed when attempting to write a Sumerian
dictionary or glossary based on these native sources.
Lexical texts (and the grammatical texts discussed immediately below) were among the
first cuneiform tablets to be found and published; their importance for the reconstruction of
Sumerian was early recognized. These texts are being systematically published in a series
entitled Materialien zum Sumerischen Lexikon (MSL). Volume 1 , edited by Benno Lands
berger, appeared in 1937; volume 1 6 appeared in 1976, and several other volumes are still
276 Appendix 2
in the process of being prepared.
The role of the lexical lists is usually described as essentially a learning device in the
schools. Mogens Trolle Larsen has discussed the "place of the lexical tradition in the
cognitive scheme ofthe ancient Mesopotamians":
There is no doubt that the lists did function within the scribal world as part of
the basic curriculum in all periods of Mesopotamian history, but it is likewise
possible to understand the lists as serving another purpose, to present a
systematic and ordered picture of the world (1987:208-209).
Westenholz says somewhat the same thing:
In many illiterate cultures, an enormously detailed vocabulary of plants,
animals, trees, etc. functions as the skeleton of an integrated classification
and taxonomy of the known world; and we may see the Sumerian lists of
everything from gods to milk-pots in a similar light as an itemized statement
on the world order, the origin and functioning of which mythology describes
in literary terms (1985:295).
Grammatical texts
The Sumerian language differs from Akkadian not only in vocabulary, but also in
morphology and syntax. These differences led the Akkadian scribes to produce a series of
grammatical texts, in addition to the lexical texts just discussed. The earliest of these are
known as the "Old Babylonian Grammatical Texts", or OBGT. Dating from about 1700-
1600 BC, these are in the form of paradigms: paradigms of verbs, nominal forms, particles,
etc. The Sumerian is on the left, and the Akkadian (if present) on the right. OBGT VI, for
example, lists over 200 forms of the verb gar, "to place". Lines 124- 125 read as follows:
mu-un-gm:
mu-gru:
This seems to be saying that for the Akkadian scribes, the forms mu-un-gar and mu
gar have the same meaning. This is how modern Sumerologists interpret the two forms, as
a difference in orthography, not in morphology.
Lines 142- 143 of the same tablet read:
ba-an-gar
ba-gar
The Akkadian scribes interpreted the difference in the Sumerian conjugation-prefix as
a difference in the Akkadian stem: mu-un-gar was translated by the B-preterite ikun, but
ba-an-gar by the Bt-preterite itakan (there is some reason to assume that itakan is a Bt
preterite, not a B-perfect).
A later series, the "Neo-Babylonian Grammatical Texts" (NBGT), dates from about
the sixth century BC. These texts are organized according to morpheme: A Sumerian
morpheme is glossed by an Akkadian equivalent. For example, NBGT I, line 153, reads:
 Sources 277

The tablet is equating the Sumerian comitative case-marker da with the Akkadian
preposition qadu, meaning "with".
Lines 405-408 of the same tablet read:

The Sumerian cohortative modal-prefix gg!, and three morphophonemic alternants of the
desiderative modal-prefixb, are all "translated" as the Akkadian desiderative-marker lu.
The NBGT texts occasionally add scribal comments or annotations, in both Sumerian
and Akkadian. Examples include the Sumerian word AN-TA, "prefix", and the Akkadian
expression a iten, "singular". These kinds of annotations do not appear in OBGT.
OBGT and NBGT are rather extensive. One might think that these texts could furnish
a key to Sumerian morphology. Unfortunately, it is not so. These texts are all relatively
late. They represent Akkadian-speakers' understanding of Sumerian. However, these
Akkadian-speaking scribes did not always understand Sumerian grammatical categories
and distinctions. For example, in the passages from OBGT cited above, a difference in
Sumerian conjugation-prefix was equated with a distinction in Akkadian stem. It is
difficult to say how accurate an equation this is. Or, Black has pointed out that OBGT V
makes a consistent distinction between the first person suffix lenl, written -en, and the
second person suffix lenl, written --en (1984:7). Black thinks that this might indicate a
difference in pronunciation, but more likely it is a purely graphic distinction.
Regarding the Sumerian and Akkadian grammatical terms which sometimes occur as
annotations, Black has also said that there is
a growing body of evidence that the scribes responsible for introducing the
grammatical terms into the grammatical analysis texts sometimes misun
derstood their meaning (or misunderstood the texts into which they were
introducing them) (1984:90).
There is also a certain amount of systematization and schematization in these tablets.
But at the same time, there is no unified method of organizing the data. As mentioned
above, OBGT has over 200 lines of gm:, but it is not always easy to follow the principles by
which these forms are organized. Other OBG T texts follow their own organization.
Like the lexical lists, there also occur mechanical errors due to problems of text
transmission. There are also scribal conventions and abbreviations. And as mentioned
earlier, there was undoubtedly a large oral component which accompanied the study of
these texts, a component which is no longer accessible. These problems (and others) mean
that although these grammatical texts are a font of useful information, this information
cannot be used uncritically. These texts cannot be viewed as an exact reflection of
Sumerian of the Ur III or earlier periods. Black has said that "in some cases it seems that
we know Sumerian better than the compilers, or copyists, of our texts" (1984:7). This is
especially true regarding the Sumerian aspectual differences, and also regarding Sumerian
278 Appendix 2
causative sentences, a type of sentence heavily dealt with in OBGT. (Similarly, Black
believes that certain Akkadian forms seen in OBGT were "especially concocted to set
against certain Sumerian forms" [ 1984:29].)
Thus, although Jacobsen is undoubtedly correct in saying that the OBGT
constitute without question the most important single group of sources both
for the history of grammatical studies generally and for our understanding of
Sumerian grammar specifically so far known ( 1 956: 1 *),
these texts must be approached with caution. Jacobsen adds that "the immensity of the
number of problems raised, and the relative insufficiency of our present knowledge of
Sumerian becomes only too clear as one approaches the texts in earnest" ( 1 956:2*).
As the above quotes exemplify, scholars vary in their estimation of the worth of these
texts in reconstructing Sumerian grammar. Jacobsen, for example, sees them as extremely
important; Krecher, on the other hand, in his study of the conjugation-prefixes containing
an fmf element, found them to be of little value ( 1985:34).
Finally, Civil has pointed out:
To my knowledge, the fundamental question: are the grammatical texts des
criptive or prescriptive ? has never been formulated ... at least in print, al
though there is widespread skepticism about their descriptive adequacy
( 1986:72).
For a long time, it was believed that OBGT texts represented mankind's first attempt to
formulate "paradigms"; these texts antedate the grammatical studies of the Indians by over a
millennium. However, what may well represent the first attempt known to organize verbal
forms into a paradigm is now found among the Ebla tablets. From a rather variegated
monolingual lexical text (TM.75.G.2260), the following lines appear:

(line 1 2) : in-na-sum
I-na-sum
nU-l-na-sum
hi-na-sum
ba-til
nu-til
in-til
hi-til

For several reasons, one might be hesitant about considering these lines to be a
"paradigm". But they do seem to indicate that the "mind" of the scribe(s) was heading in
that direction. And this tablet antedates the OBGT texts by some 500 years.
The function of the lexical texts (and the smattering of what might be called a
"grammatical text" just discussed) at Ebla was similar to the function of the lexical and
grammatical texts found throughout Mesopotamia and the Ancient Near East in general: to
enable scribes to master Sumerian. For the Eblaite scribes, however, Sumerian was a
language still being spoken. For the Akkadian scribes of the Old Babylonian period,
Sumerian had virtually ceased to exist as a spoken language, and was only a language of
the schools.
 Sources 279

OBGT and NBGT were published in MSL IV (1956). These texts were prefaced by
Jacobsen with a discussion of the Sumerian verbal system as reflected in these texts.
Jacobsen has also written a very interesting article intended for a more general linguistic
audience, discussing the system of paradigms seen in OBGT and NBGT (1974). Black
(1984) has written a book especially on these grammatical texts, and on the philosophy of
language which they represent; this work also has much incidental discussion of various
aspects of Sumerian morphology.
Very recently, Civil et al have published some "Middle Babylonian Grammatical
Texts" ( 1986). These have not yet been fully studied.
Syllabic S umerian
In addition to the "standard" or "normal" orthography and spelling of Sumerian, there
is a certain amount of what is called "syllabic Sumerian" or "phonetic Sumerian". Standard
Sumerian is written using a combination of logographic and syllabic signs. Syllabic
Sumerian, however, is written using only syllabic signs. For example, the standard Su
merian orthography for a locative phrase, kalam-ma, is written in syllabic Sumerian as ka
la-ma.
There are not a great deal of texts in syllabic Sumerian; they are all relatively late.
Interestingly, not many syllabic texts come from the Mesopotamian heartland; they are
mostly from northern Babylonia, or farther afield. The practice probably originated in the
scribal school system as a device for the scribes to cope with the difficulties of standard
Sumerian orthography'. In the case of syllabically-written incantations and liturgical texts,
the purpose was probably to aid in correct recitation.
Unlike the lexical and grammatical texts discussed above, there is no standard or
canonical system of writing syllabic Sumerian; it varies to some degree from text to text.
Much of it was probably produced on an ad-hoc basis, to deal with particular texts.
Since syllabic Sumerian is an attempt to reproduce spoken Sumerian, it should reveal
some of the features not shown in the normal orthography. Thus, one might think of it as
another key to unlocking Sumerian morphology. Unfortunately, it is very difficult to
understand syllabic Sumerian, even more difficult than it is to understand Sumerian in
standard orthography. The reason is precisely because standard Sumerian masks certain
phonetic problems, such as morphophonemic alternation, contraction, assimilation, etc.
When such phenomena actually show up in syllabic Sumerian, it is often difficult to
untangle the forms. Even in cases where the same text is preserved once in standard
orthography and once in syllabic orthography, the phonetic relationship between the two is
not always easy to see.
A relatively simple case is the writing at-ta, for standard an-ta, "from the sky", or
"from above". Should it be assumed that the standard Sumerian was also pronounced
latta/, and that the written form an-ta is a morphographemic or historical writing? If so,
should the an-sign be transliterated by an at-value? Perhaps in early Sumerian, the word
was indeed pronounced lanta/, but an assimilation took place in later Sumerian, producing
latta/. How can this change be dated? On the other hand, perhaps such a writing as at-ta
reflects the Akkadian assimilation of nasals, and doesn't say anything about Sumerian.
280 Appendix 2
Even in this one simple instance, one can think of several variables which must be
taken into account. But consider an even more complicated case. From the root hlL
meaning "to be confused", there appears a man1 form in standard orthography as ba-bir
bir-re (this shows formation of the marii by means of both reduplication and the marii-suf
fix .e). This appears in syllabic orthography as ba-bi-ib- presumably representing
Ibabibre/. Does this mean that the writing in standard orthography, ba-bir-bir-re, should
also be understood as representing Ibabibre/, and that this is a morphographemic or
historical spelling? How should it be transliterated?
Such examples illustrate the extent to which phonetic processes are masked by
standard orthography, and they show the difficulty in interpreting the syllabic forms. And
since syllabic Sumerian varies to some degree from text to text, it is difficult to generalize
about what is seen. All syllabic texts are rather late, from the Old B abylonian period or
after. They thus reflect a stage when Sumerian was no longer a spoken language, so to
some degree the phonetic differences that appear may be conditioned by the Akkadian
language of the scribes.
Very recently, however, there have been found at Ebla syllabically-written versions of
lexical lists. Civil has called their existence "a most unexpected surprise which opens a
new chapter in the understanding of the earliest lexical compilations and provides
phonological data for the oldest stages of Sumerian" ( 1 982: 1 ) . These syllabically-written
lexical texts are very difficult to interpret, and only preliminary work has been accom
plished.
Much more work, in general, remains to be done on syllabic Sumerian. The two
examples given above show the kinds of information which such texts can provide. A
more thorough investigation might help in solving some of the perplexing problems
encountered in the Sumerian writing system.
The Emesal dialect of Sumerian is written in a mixture of standard (i.e. , mostly
logographic) Sumerian and of syllabic Sumerian. For example, the word for "lady" in
Emesal is gaan, corresponding to Main Dialect nin. Sometimes Emesal texts simply use
the same nin-sign; it is assumed that the "reader" will know enough to render the nin-sign
as the Emesal equivalent gaan. More frequently, however (at least, with this particular
word), the word is spelled syllabically, gf!-a-an.
An Emesal vocabulary has also been preserved, some 1 77 lines long. It gives the
Emesal form in the far left-hand column; the Main Dialect form; and an Akkadian
translation in the far-right column. A typical example is line 96:

This tells us that the adjective "wide", Akkadian rapu, which is written with the dagal-sign
in Main Dialect, is I damall in Emesal.

Bilinguals
A certain number of "bilingual" texts have come down to us. These are of two main
types. In "interlinear" texts, a line of Sumerian is followed by a line of Akkadian. There
are many incantations of this type. Other texts are written in "parallel columns", with the
 Sources 28 1

Sumerian on the left of a tablet, and the Akkadian on the right.

Bilingual texts are of many different genres. They include incantations, rituals, hymns,
proverbs, letters, and even a few royal inscriptions. Sometimes even more than two
languages are involved. For example, among the texts found at Boghazkoy (a Hittite
speaking area), there is a trilingual poem to the god Ishkur. The text is divided into groups
of four lines. The first line is Sumerian in standard orthography; the second line is the
same in syllabic Sumerian; the third line is an Akkadian translation; the fourth line is a
Hittite translation. This text illustrates how syllabic Sumerian was used to help the Hittite
speaking scribes in trying to figure out the standard Sumerian version.
Most bilinguals are rather late, from the later Old Babylonian period on. Some may be
earlier; for example, there are a couple of Old Babylonian copies of Old Akkadian bilingual
royal inscriptions. One is a bilingual inscription of Sargon, written in parallel columns on
the back of a statue of Sargon; this was re-copied in Old Babylonian times.
Such terms as "bilingual" or "trilingual" are mostly used to refer to single tablets
which have writing in more than one language. There are other cases where a Sumerian
version of a text is present on one tablet, and an Akkadian translation on another tablet.
(Without both versions, it might never be known that the Akkadian was a translation.) The
term "bilingual" is occasionally used to refer to such "parallel versions".
One of Shulgi's royal inscriptions has been preserved in both a Sumerian version and
an Akkadian version. But neither appears to be a translation of the other; rather, they seem
to be independent compositions commemorating the same event, the building of a temple.
Such texts (there are few) are helpful in elucidating the relationship between Sumerian and
Akkadian during the Ur III period.
Other sources
There are a few other sources to aid in reconstructing Sumerian. For instance, there
are Akkadian commentaries to certain genres of text, such as medical texts and omina.
None of these are as extensive or as helpful as the four groups outlined above.
 Appendix 3
Glossary

There is no standard sign-list for Sumerian. However, the sign-lists of Borger and
Labat, even though based on Akkadian values, are still useful for Sumerian. Labat is
especially helpful for the study of the palaeography of the signs, that is, the variation in
their basic shape throughout time. Borger provides some information about Sumerian
grammar, and also a certain amount of bibliographic material. The two volumes of
Ellermeier which have appeared to date are useful in sorting out inconsistencies in
published transliterations.
At the moment, there is no up-to-date dictionary of Sumerian. For a number of years,
the University of Pennsylvania has been preparing just such a project, the Pennsylvania
S umerian Dictionary (PSD; cited in the Bibliography under Sjoberg [ 1 984]). As of this
writing, the only volume which has appeared so far is for the letter "B". The glossary of
Delitzsch, although dating to 1 9 14, is still one of the most useful single-volume dic
tionaries, although not the easiest to use. There are a few glossaries to specialized bodies of
texts. One of the more recent is Behrens and Steible ( 1983). Unfortunately, even semi
serious lexicographical work usually means looking at the glossaries and indices of many
different text editions. For the beginning student, perhaps the most useful of these is
Sollberger ( 1 966). Professional Sumerologists keep very large and detailed files on
S umerian words. The core of the PSD, for example, is Ake Sjoberg's collection of over
500,000 dictionary entries, which he started in 1 949.

a (Text 1 )
arm, strength (Text 22)
a ... fU to dedicate a votive object (Text 4)
ab (Text 17)
abzu apsu (Text 14)
ad (Text 1 9)
ad-da father (Text 1 9)
Adab Adab (GN) (Text 20)
g (Text 8)
i!gi!2 (Text 8)
alam statue (Text 1 5)
am (Text 1 1 )
ama mother (Text 1 5)
amar young bull (Text 1 3)
d
Amar- Zuen Amar-Sin (PN) (Text 1 3)
an (Text 1 8)
An An (DN) (Text 6)
an heaven (Text 1 )

283
284 Appendix 3

An-nu-ni-tum Annunitum (DN) (Text 1 8)

an-ub corner (Text 10)
arad servant (Text 20)
a (Text 22a)
ba (Text 10)
ba to give as a gift (Text 21c)
ba6 (Text 1 1)
ba-al to dredge (Text 5)
Ba-ba6 Baba (DN) (Text 1 1)
Ba-ba6-nin-am Babaninam (PN) (Text 1 1)
bad city wall, rampart, fortification (Text 3)
barag (bara dais (Text 6; Discussion, Lesson 15)
barag-sig9-g pedestal (Text 15)
be (Text 9)
bi (Text 9)
bi (Text 1 2)
bi! (Text 1 2)
Bil-g-mes Gilgamesh (DN) (Text 12)
bU6 to tear out, to uproot (Text 1 5)
da (Text 1 2)
da side (Text 10; Discussion, Lesson 12)
Da-da Dada (PN) (Text 21b)
dab S to hold (Text 1 1 )
dadag to be clean, pure (Text 19)
dam wife, consort (Text 1 8)
digi!: god (Text 6)
dim (Text 1 2)
dim to fashion, fom1 (Text 1 1 )
du to build (Text 1; Text 1 1)
dub tablet (Text 19a)
dub-sar scribe (Text 19a; Discussion, Lesson 2 1b)
dumu son (Text 9)
dur bond (Text 1 2a)
(Text 12)
house (Text 1)
e (Text 22)
E-dur-an-ki Eduranki (TN) (Text 1 2a)
.e-kur Ekur (TN) (Text 8)
.e-ag4-g-pad-da Eshagepada (TN) (Text 19)
E-temen-ni -guru3 Etemenniguru (TN) (Text 9)
eb (Text 15)
en (Text 2)
en lord (Text 9)
 Glossary 28 5

en priest, priestess (Text 17)

En-dim-gig Endimgig (ON) (Text 1 2)
En-erinTnun Enerinnun (ON) (Text 5)
En-ki Enki (DN) (Text 14)
En-lil Enlil (DN) (Text 5)
ensi2 city governor, local ruler (Text 20)
er (Text 22a)
erin2 (Text 5)
g (Text 2)
gal great (Text 19; Discussion, Lesson 7)
g (Text 1 6)
g (Text 20)
gi (Text 2)
g to return, to restore (Text 9)
g!-na true, correct; standard, certified (Text 21a)
gi-na to standardize, to certify (Text 2 1d)
gig (Text 1 2)
gg (Text 22)
gg-za throne (Text 22)
gg-za-Ia (kind of official) (Text 22)
gub to stand; to make stand, to plant (Text 6)
gudug (gudu4 ) (kind of priest) (Text 19; Text 1 1)
guru3 (Text 9)
M (Text 8)
gal (Text 19)
getin vine, wine (Text 1 6)
getug3 ear, intelligence (Text 22)
gi6 (Text 17)
g!6-Qill: giparu (Text 17)
gir (Text 1 1)
gg (determinative preceding objects of wood) (Text 6)
b (Text 1 2)
Ij-a-b-me-er Hashamer (PN) (Text 22a)
Ij-ba-Iu 5-g Habaluge (PN) (Text 20)
Ij-la-dBa-ba6 Halababa (PN) (Text 19a)
b (Text 1 5)
bi (Text 1 1)
bi-li attraction; headdress, wig (Text 1 1)
! (Text 2 1 b)
i oil, fat (Text 1 6)
I-bi-dZuen Ibbi-Sin (PN) (Text 21b)
i-nun butter (Text 1 6)
ib (Text 1 2)
2 86 Appendix 3

id (Text 1 9)
id (h) river, canal (Text 5)
il (Text 22)
im (Text 1 5)
in (Text 7)
Inanna Inanna (DN) (Text 2)
iq (Text 1 9)
ir (Text 1 4)
is (Text 22a)
IS-ku-un-dZuen Ishkun-Sin (GN) (Text 22a)
isib (kind of priest) (Text 1 9)
ka (Text 5)
kalag (kala) to be mighty (Text 2)
kalam land (Text 17)
kar quay, pier; market place (Text 17)
Kar-zid-da Karzida (GN) (Text 17)
ke4 (Text 1 ; Discussion, Lesson 1 )
ki (determinative following GNs) (Text 1 )
ki place, eanh (Text 6; Text 1 2a; Text 1 4)
g
ki ... 2 to love (Text 8)
Ki-en-gi Sumer (GN) (Text 2)
Ki-lul-la Kilula (PN) (Text 22)
Ki-uri Akkad (GN) (Text 2)
kiri6 garden (Text 6)
kisib cylinder-seal (Text 22)
ku (Text 22a)
kug Gill) bright, pure, holy (Text 1 7 ; Discussion, Lesson 1 8)
kur mountain; highland; foreign land (Text 5)
kur to change (Text 1 5)
kurs Ck!! s ) (Text 1 2)
kur9 (ku4 ) to enter (Text 17)
la (Text 4)
la (Text 7)
la to hold, to lift, to carry (Text 22)
Lagas Lagash (GN) (Text 22)
lal honey (Text 1 6)
lam (Text 22)
Lamar DN (Text 1 1 )
le (Text 1 3)
li (Text 1 1 )
li (Text 1 7)
lil air, wind (Text 5)
limmu2 four (Text 1 0)
 Glossary 287

lil man (Text 7)

Iu s (Text 20)
lugal king (Text 1 ; Discussion, Lesson 7)
lul (Text 22)
ma (Text 1 )
ma-da land (Text 1 9)
ma-na mina (Text 2 1 a)
man to be splendid, magnificent (Text 6)
mar (Text 1 9)
Mar-tu the Amorites; the West (Text 1 9)
me (Text 1 5)
mes (Text 22)
Mes-lam-ta-- Meslamtaea (DN) (Text 22)
me (Text 1 2)
mu (Text 1 )
mu name (Text 1 5)
mu ...pad (Illi) to propose (Text 1 3)
Mu-ri-iq-Ti-id-ni-im Muriq-Tidnim (ON) (Text 1 9)
mu-sar-ra inscnptIon (Text 1 2)
munus woman (Text 1 1 )
na (Text 1 )
nam (Text 4; Discussion, Lesson 9; Sign-list, Lesson 1 2)
nam ... kurS (ku S ) to curse (Text 1 2)
Nammu Nammu (DN) (Text 1 )
Nanibgal Nanibgal (DN) (Text 2 1 b)
Nanna Nanna (DN) (Text 1 )
Nane Nanshe (DN) (Text 1 1 )
ne (Text 6)
ne forces, troops (Text 1 9)
m (Text 1 )
ni (Text 9)
Nibru Nippur (ON) (Text 1 3)
nidba food offering (Text 5)
ni g Crrl) thing (Text 22; Discussion, Lesson 4; Discussion, Lesson 9)
nin lady, mistress; 'lord' (Text 1; Discussion, Lesson 1 1 )
Nin-gal Ningal (DN) (Text 4)
Nin-gir-su Ningirsu (DN) (Text 1 1 )
Nin-li l Ninlil (DN) (Text 1 9)
Nin-tur-dumu-g1! l Q Ninturdumugu (PN?) (Text 1 6a)
nir (Text 1 9)
nir-gal prince (Text 1 9)
nitan (nita) man, male (Text 2)
nu (Text 1 8)
2 88 Appendix 3

nu not (Text 1 6)
numun seed; offspring, progeny (Text 1 5)
nun (Text 5)
nun prince, noble (Text 1 6)
pad (w) to find, call, reveal (Text 1 3)
nill: (Text 1 7)
ra (Text 5)
re (Text 6)
re6 (Text 1 6a)
ri (Text 1 9)
ru to send (Commentary, Lesson 4)
sag head (Text 9)
sag slave (Text 2 1 c)
Sag- dNanna-zu Sagnannazu (PN) (Text 2 1 c)
sag-lis supporter, sustainer, patron (Text 1 3)
sagi! (kind of priest) (Text 2 1 c)
sar (Text 1 2)
sar to write (Text 1 9a)
sig9 (si) to be narrow (Text 1 5)
sikil to be pure, clean (Text 6)
sipad (sipa) shepherd (Text 1 9)
siskur2 sacrifice (Text 1 6)
su (Text 1 1 )
sud4 13) to be long (Text 1 7)
sag4 (sa.) heart (Text 1 8)
sags (g6 ) to be good (Text 22)
S ara
2 Shara (DN) (Text 1 9)
se (Text 4)
silig to cease (Text 1 6)
su (Text 1 8)
su hand (Text 1 9)
su ... ur to erase (Text 1 2)
d
S u- Zuen Shu-Sin (PN) (Text 1 8)
suI (Text 1 0)
S ul-gi Shulgi (PN) (Text 1 0)
ta (Text 1 7)
temen (Text 9)
ti (Text 1 9)
til to put an end to (Text 1 5)
til Cti) to live (Text 4; Commentary, Text 22)
tu (Text 1 9)
turn (Text 1 8)
tur (Text 1 6a)
 Glossary 289

!! (Text 1 1 )
!! and (Text 1 6)
ub (Text 1 0)
ud ) day (Text 1 2)
ul remote, distant (Text 1 7)
un (Text 17)
Unug (Unu) Uruk (ON) (Text 9)
ur man, warrior (Text 1 )
Ur-ba- ? Urba? (PN) (Text 22)
d
Ur- Lamar Ur-Lamar (PN) (Text 1 9a)
d
Ur_ Nammu Ur-Nammu (PN) (Text 1 )
Ur-dNanibgal Ur-Nanibgal (PN) (Text 2 1 b)
Ur-dNin-grr-su Ur-Ningirsu (PN) (Text 1 1 )
ur (Text 1 2)
Uri (Text 2)
Urims (Uri s ) Ur (ON) (Text 1 )
us (Text 1 3)
utu sun (Text 17)
Utu Utu (DN) (Text 4a)
zabar bronze (Text 1 1 ; Discussion, Lesson 20)
zabar-dab S (kind of official) (Text 1 1 )
zid (zi) effective, true (Text 1 7)
zu (Text 1 3)
zu your (Text 2 1 b)
zu to know (Text 2 1 c)
Zuen Zuen (DN) (Text 1 3 )
"5" (Text 2 1 a)
 Appendix 4
Bibliography

Abbreviations

AfO Archiv far Orientforschung

AHw Akkadisches Handworterbuch
AIUON Istituto Universitario Orientale di Napoli - Annali
AJA American Journal of Archaeology
AJSL American Journal of Semitic Languages
ArOr Archiv Orientalia
AS Assyriological S tudies
ASJ Acta S umerologica (Japan)
BiOr B i bliotheca Orientalis
CAD Chicago A ssyrian Dictionary
CRAI Compte rendu de la Rencontre Assyriologique Internationale
GLECS Groupe linguistique d'etudes chamito-semitiques - comptes rendus
HUCA Hebrew Union College Annual
JANES Journal of the Ancient Near Eastern Society
JAOS Journal of the American Oriental Society
JCS Journal of CuneifOlID Studies
JNES Journal of Near Eastern Studies
MIO Mitteilungen des Instituts fur Orientforschung
MSL Materialien zum S umerischen Lexikon
OA Oriens Antiquus
OIP Oriental Institute Publications
Or Orientalia
PSD Pennsylvania Sumerian Dictionary
RA Revue d'Assyriologie
RLA Reallexikon der Assyriologie
SEb Studi Eblaiti
SEL Studi Epigrafici e Linguistici
TCS Texts from Cuneiform Sources
UF U garit-Forschungen
VO Vicino Oriente
WO Wel t des Orien ts
WZKM Wiener Zeitschrift far die Kunde des Morgenlandes
ZA Zeitschrift fOr Assyriologie

Works Cited

van Aalderen, C.T. 1 982. "Some observations on ergativity and Sumerian". Orientalia
Lovaniensia Periodia 1 3 : 25-44.

29 1
292 Appendix 4

Attinger, Pascal. 1985. "Les prefixes absolutifs de la premiere et de la deuxieme personne

singulier dans les formes maru ergatives". ZA 75: 161- 178.
Bauer, Josef. 1975-76. "Zur /dr/-Phonem des Sumerischen". W0 8: 1 -9.
Behrens, Hermann and Horst Steible. 1983. Glossar zu den altsumerischen Bau- und
Weihinschriften. Wiesbaden: Franz Steiner Verlag. Freiburger Altorientalische Stu
dien 6.
Biga, Maria Giovanna. 1980. "Bilinguismo sumerico-accadico nell'epistolografia sumerica
e paleoaccadica". V0 3: 121-129.
Biggs, Robert D. 1974. Inscriptions from Tell Abu Salabikh. Chicago: University of
Chicago Press. OIP 99.
Black, Jeremy. 1984. Sumerian Grammar in Babylonian Theory. Rome: Pontifical Bib
lical Institute. Studia Pohl, Series maior 12.
Boisson, Claude. 1988. Constraintes typologiques sur le systeme phonologique du Su
merien (to appear).
Borger, Rykle. 1978. Assyrisch-babylonische Zeichenliste. Kevelaer: Verlag Butzon &
Bercker. Alter Orient und Altes Testament 33.
Butz, K. 1973-74. "Konzentrationen wirtschaftlicher Macht im Konigreich Larsa: Der
Nanna-Ningal-Tempelkomplex in Ur". WZKM 65-66: 1-58.
Canby, Jeanny Vorys. 1987. "A Monumental Puzzle: Reconstructing the Ur-Nammu
Stela". Expedition 29/ 1 : 54-64.
Christian, Viktor. 1961. Die Herkunft der Sumerer. Vienna.
Civil, Miguel. 1973a. "The Sumerian Writing System". Or42: 21-34.
1973b. "From Enki 's Headaches to Phonology". JNES 32: 57-61 .
1976. "Lexicography". AS 20: 123-157.
1982. "Studies on Early Dynastic Lexicography. I". OA 21: 1 -26.
1985. "On Some Texts Mentioning Ur-Namma". Or 54: 27-45.
Civil, Miguel, Oliver R. Gurney, Douglas A. Kennedy. 1986. Materials for the Sumerian
Lexicon: Supplementruy Series 1 . Rome: Pontifical Biblical Institute.
Clauson, Sir Gerard. 1973. "Philology and Archaeology". Antiquity 47: 37-42.
Cooper, Jerrold S. 1973. "Sumerian and Akkadian in Sumer and Akkad". Or 42: 239-
246.
Dandamayev, Muhammad A. 1982. "The Social Position of Neo-Babylonian Scribes".
in: Klengel, Horst (ed), Gesellschaft und Kultur im alten Vorderasien. Berlin: Aka
demie-Verlag. 35-39.
Delitzsch, Friedrich. 1914. Sumerisches Glossar. Leipzig. Reprint: Leipzig: Zentral
antiquariat der deutschen demokratischen Republik. 1979.
Diakonoff, I.M. 1965. Semito-Hamitic Langlli!ges: An Essay in Classification. Moscow:
Nauka.
----- 1976. "Ancient Writing and Ancient Written Language: Pitfalls and Peculiarities in
the Study of Sumerian". AS 20: 99-121.
----- 1983. "Some Reflections on Numerals in Sumerian: Towards a History of Mathe
matical Speculation". JAOS 103: 83-93.
Du Bois, John W. 1987. "The discourse basis of ergativity". Language 63: 805-855.
 Bibliography 293

Dunham, Sally. 1986. "Sumerian Words for Foundation". RA 80: 3 1-64.

Edzard, Dietz Otto. 197 1f. "ljamt!!, mara und freie Reduplikation beim sumerischen Ver
bum". ZA 6 1 ( 1 971): 208-232; ZA 62 (1972): 1-34; ZA 66 (1976): 45-61 .
----- 1975. "Zur 'Wortbildung im Sumerischen"'. ZA 65: 254-257.
----- 1976. "'Du hast mir gegeben', 'ich habe dir gegeben': Ober das sumerische Verbum
sum". W0 8: 159- 177.
----- 1976-80. "Keilschrift". RLA 5: 544-568.
----- 1980. "Sumerisch 1 bis 10 in Ebla". SEb 1 : 121-127.
Ellermeier, Friedrich. 1979f. Sumerisches Glossar. Band 1, Teil 1 : Sumerische Lautwer
te (2 vols). Gottingen: Selbstverlag Dr. Friedrich Ellermeier.
Ellis, Richard. 1968. Foundation Deposits in Ancient Mesopotamia. New Haven: Yale
University Press.
Fales, F.M. 1984. "A Survey of Two-word Eblaic Entries in the Bilingual Vocabularies".
in: Pelio Fronzaroli (ed), Studies on the Language of Ebla. Florence. Quademi di Se
mitistica 13. 173- 1 87.
Falkenstein, Adam. 1959. Das Sumerische. Leiden: E.J. Brill. Handbuch der Orien
talistik.
----- 1978. Grarnrnatik der Sprache Gudeas von Laga2 . Rome: Pontifical Biblical Insti
tute. AnalectaOrientalia 28, 29, 30.
Finegan, Jack. 1 979. Archaeological History of the Ancient Middle East. Boulder: West
view Press.
Finkel, Irving L. 1987. "An Issue of Weights from the Reign of Amar-Sin". ZA 77:
192- 193.
Foster, Benjamin. 1982. "Ethnicity and Onomastics in Sargonic Mesopotamia". Or 5 1 :
297-354.
Foxvog, Daniel. 1975. "The Sumerian Ergative Construction". Or 44: 395-425.
Frayne, Douglas. 1983. "Sulgi, the Runner". JAOS 103: 739-748.
Gadd, C.J. 197 1 . "Babylonia c. 2120-1 800 B.C." in: LE.S. Edwards et al (eds), The
Cambridge Ancient History3 . Cambridge: Cambridge University Press. Vol. 1 , Part
2. 595-643.
Ge1b, 1.J. 1948. "A New Clay-Nail of Hammurabi". JNES 7: 267-27 1 .
----- 1956. "The Names of Ex-Voto Objects in Ancient Mesopotamia". Names 4 : 65-
69.
----- 1962. "Ethnic Reconstruction and Onomastic Evidence". Names 1 0: 45-52.
----- 1977. "Typology of Mesopotamian Seal Inscriptions". in: Gibson and Biggs ( 1977)
1 09- 121.
----- 1979a. "Definition and Discussion of Slavery and Serfdom". UF 1 1 : 283-297.
----- 1979b. "Household and Family in Early Mesopotamia". in: Edward Lipinski (ed),
State and Temple Economy in the Ancient Near East. Leuven. Vol I: 1-97.
----- 1982. "Terms for Slaves in Ancient Mesopotamia". in: M.A. Dandamayev (ed),
Societies and Languages of the Ancient Near East: Studies in Honor of LM. Diako
, noff. Warrni nster: Aris and Phillips. 8 1-97.
----- 1987. "Compound Divine Names in the Ur lIT Period". in: Francesca Rochberg-
294 Appendix 4

Halton (ed), Langua,ge, Literature, and History: Philological and Historical Studies
Presented to Erica Reiner. New Haven: American Oriental Society. American Oriental
Series 67. 1 25- 1 38.
Gibson, McGuire and Robert D. Biggs. 1977. Seals and Sealing in the Ancient Near East.
Malibu: Undena. Bibliotheca Mesopotamica 6.
-----. 1987. The Organization of Power: Aspects of Bureacracy in the Ancient Near East.
Chicago: University of Chicago Press. Studies in Ancient Oriental Civilizations 46.
Gordon, Cyrus. 1987. "Eblaitica". in: Cyrus Gordon et al (eds), Eblaitica: Essays on the
Ebla Archives and Eblaite Languag. Winona Lake: Eisenbrauns. Vol. 1 : 19-28.
Gragg, Gene. 1968. "The Syntax of the Copula in Sumerian". in: I.W.M. Verhaar (ed),
The Verb "Be" and its Synonyms. Dordrecht. Vol. 3: 86-109.
----- 1972a. "Sumerian and Selected Afro-Asiatic Languages". in: P. Peranteau (ed), The
Chicago Which Hunt. Chicago: Chicago Linguistic Society. 153- 1 69.
----- 1972b. "Observations on Grammatical Variation in Sumerian Literary Texts".
JAOS 92: 204-21 3.
----- 1973a. Sumerian Dimensional Infixes. Kevelaer: Verlag Butzon & Bercker. Alter
Orient und Altes Testament, Sonderreihe 5.
----- 1973b. "A Class of 'When' Clauses in Sumerian". JNES 32: 124- 134.
----- 1973c. "Linguistics, Method, and Extinct Languages: The Case of Sumerian". Or
42: 78-96.
Grayson, A. Kirk. 1980. "Histories and Historians of the Ancient Near East: Assyria and
Babylonia". Or 49: 140- 194.
Green, M.W. 198 1 . "The Construction and Implementation of the Cuneiform Writing
System". Visible Language 15: 345-372.
Green, M.W. and Hans I. Nissen. 1987. Zeichenliste der archaischen Texte aus Uruk.
Berlin: Gebr. Mann Verlag. Archaische Texte aus Uruk Band 2.
Haayer, G. 1 986. "Languages in Contact: The Case of Akkadian and Sumerian". in:
H.L.I. Vanstiphout et al (eds), Scripta Signa Vocis. Groningen: Egbert Forsten. 77-
84.
Hallo, William W. 1957. Early Mesopotamian Royal Titles: A Philologic and Historical
Analysis. New Haven: American Oriental Society.
1962. "The Royal Inscriptions of Ur: A Typology". HUCA 33: 1-43.
----- 1963. "Lexical Notes on the Neo-Sumerian Metal Industry". BiOr20: 1 36-141.
----- 1966. "The Coronation of Ur-Nammu". JCS 20: 133-141.
Hallo, William W. and 1.1.A. van Dijk. 1968. The Exaltation of Inanna. New Haven:
Yale University Press. Yale Near Eastern Researches 3.
Heimpel, Wolfgang. 1987. "Gudea's Fated Brick". JNES 46: 205-2 1 1 .
Horsnell, Malcolm. 1977. "The Grammar and Syntax of the Year-Names of the First Dy
nasty of Babylon". JNES 36: 277-285.
lacobsen, Thorkild. 1953. "The Reign of Ibbi-Suen". Reprinted in: Moran (1970) 173-
186; 408-41 6.
----- 1956. "Introduction to the Chicago Grammatical Texts". in: MSL IV, 1 *-50*.
 Bibliography 295

----- 1957. "Early Political Development in Mesopotamia". Reprinted in: Moran (1970)
1 32- 156; 366-396.
1965. "About the Sumerian Verb". Reprinted in: Moran (1970) 245-270; 430-466.
1968. "Some Sumerian City-Names". JCS 2 1 : 100- 103.
1973. "Notes on the Sumerian Genitive". JNES 32: 161-166.
1974. "Very Ancient Texts: Babylonian Grammatical Texts". in: D. Hymes (ed),
Studies in the History of Linguistics: Traditions and Paradigms. Bloomington:
Indiana University Press. 41-62.
----- 1987. The Harps that Once... Sumerian Poetry in Translation. New Haven: Yale
University Press.
----- 1988. "Sumerian Grammar Today". JADS 108: 123- 1 33.
Kaneva, I.T. 1970. "Participles in Sumerian". MID 16: 541-565.
Kang, S.T. 1968. A Study of the Sumerian Verb in Bilingual Grammatical Texts. Cin-
cinnati: Hebrew Union College [dissertation].
Kienast, Burkhart. 1975. "Zur Wortbildung des Sumerischen". ZA 65: 1-27.
----- 198 1a. "1st das Neusumerische eine lebende Sprache?" CRAI28: 105- 1 1 1 .
----- 198 1 b. "Probleme der sumerischen Grammatik". ZA 70: 1-35.
Kramer, Samuel Noah. 1936. "Studies in Sumerian Phonetics". ArOr 8: 1 8-33.
----- 1940. Lamentation over the Destruction of Ur. Chicago: University of Chicago
Press. AS 1 2.
----- 1967. "The Death of Ur-Nammu and His Descent to the Netherworld". JCS 2 1 :
104-122.
----- 1983. "The Ur-Nammu Law Code: Who Was Its Author?" Or 52: 453-456.
Kraus, ER. 1958. "DLtil.1a. Sumerische Prozessprotokolle und Verwandtes aus der Zeit
der Ill. Dynastie von Ur". BiOI 15: 70-84.
Krecher, Joachim. 1967. "Zum Emesal-Dialekt des Sumerischen". in: Heidelberger Stu
dien zum Alten Orient. Wiesbaden: Otto Harrassowitz. 87- 1 1 O.
----- 1967-68. "Die pluralischen Verba fur 'gehen' und 'stehen' im Sumerischen". WD 4:
1-1 1.
----- 1978. "Die Form und der Gebrauch der nominalen Verbalformen und die Determina
tion im Sumerischen". Dr47: 376-403.
1985. "Die Im/-Prilfixe des sumerischen Verbums". Or 54: 1 33- 1 8 1 .
----- 1987a. "DU KUx (-r) 'eintreten', 'eineinbringen"' . ZA 77: 7-2 1 .
=

----- 1987b. "Morphemeless Syntax in Sumerian as Seen on the Background of Word-

composition in Chukchee". ASJ9: 67-88.
Krispijn, Th. 1982. "Das Verb in den zweisprachigen Inschriften der Hammurabi
Dynasty". in: G. van Driehl (ed), Zikir umim: Assyriological Studies Presented to
ER. Kraus. 145-162.
Kupper, J.-R. and Edmond Sollberger. 197 1 . Inscriptions royales sumeriennes et akka
diennes. Paris.
Kutscher, Eduard Yechezkiel. 1982. A History of the Hebrew Lang!illg. Leiden: E.J.
Brill.
Kutscher, Raphael. 1975. Oh Angry Sea (a-ab-ba hu-Iuh-ha): The History of a Sumerian
296 Appendix 4

Cong@gational Lament. New Haven: Yale University Press. Yale Near Eastern Re
searches 6.
Labat, Rene. 1 976. Manuel d'epigraphie akkadienne5 . Paris: Paul Geuthner.
Lambert, Maurice. 1 957. "Les noms du pere en Sumerien". Proceedings of the Twenty
Second Congress of Orientalists. Leiden: E.J. Brill. Vol. 2: 27-29.
----- 1 972-7 3 . "Probleme essentiel dans l'etude du Sumerien: Les prefixes verbaux".
GLECS 1 7 : 1 3-2 1 ; 97- 1 04.
Lambert, W.G. 1 975. (review of Henri Limet, Les legendes des sceaux cassites [ 1 97 1 ] ) .
BiOr 32: 2 1 9-223.
Larsen, Mogens Trolle. 1987. "The Mesopotamian Lukewarm Mind: Reflections on
Science, Divination, and Literacy". in: Francesca Rochberg-Halton (ed), Lang.!!M
Literature, and History: Philological and Historical S tudies Presented to Erica Reiner.
New Haven: American Oriental Society. American Oriental Series 67. 203-205.
Lieberman, Stephen J. 1 977. The Sumerian Loanwords in Old-Babylonian Akkadian.
Missoula: Scholars Press. Harvard Semitic S tudies 22.
----- 1 978. (letter to editor). Scientific American 239/ 1 0: 1 5 .
----- 1 979. "The phoneme /0/ in Sumerian". in: Studies in Honor of Tom B. Jones.
Kevelaer: Verlag Butzon & Bercker. Alter Orient und Altes Testament 203 . 2 1 -28.
----- 1 980. "Of Clay Pebbles, Hollow Clay Balls, and Writing: A Sumerian View".
AlA 84: 339-358.
Limet, Henri. 1 975a. "Le morpheme suffixe /-a/ en Sumerien". RA 69: 5- 1 8 .
----- 1 975b. "Permanence et changement dans l a toponymie de l a Mesopotamie antique".
in: La Toponymie antique. Actes du Colloque de Strasbourg 1 2- 1 4 juin 1 975. Leiden:
EJ. Brill. 83- 1 1 5 .
----- 1 978. "Etude semantique d e ma.da, kur, kalam". RA 7 2 : 1 - 1 2.
Martiny, Giinther. 1 940. "The Orientation of the Gimilsin Temple and the Palace Chapel".
in: Frankfort, Henri, Seton Lloyd, and Thorkild J acobsen, The Gimilsin Temple and
the Palace of the Rulers at Tell Asmar. Chicago: University of Chicago Press. OIP
43. 92-96.
Michalowski, Piotr. 1 980a. "Sumerian as an Ergative Language". lCS 32: 86- 1 03.
----- 1 980b. "Konigsbriefe". RLA 6: 5 1 -59.
----- 1 987. "Charisma and Control: On Continuity and Change in Early Mesopotamian
Bureacratic S ystems". in: Gibson and Biggs ( 1 987) 50-68.
Moorey, P.R.S . 1984. "Where did they bury the kings of the IIIrd dynasty of Ur?". Iraq
46: 1 - 1 8.
Moran, William (ed). 1 970. Toward the Image of Tammuz. Cambridge: Harvard Univer
sity Press.
Nissen, Hans J. 1 985. "The Emergence of Writing in the Ancient Near East". Inter
disciplinary Science Revie ws 1 0/4: 349-36 1 .
----- 1 986. "The archaic texts from Uruk". World Archaeology 1 7/ 3 : 3 1 7-334.
----- 1988. The Early History of the Ancient Near East: 9000-2000 B .C. Translated by
Elizabeth Lutzeier, with Kenneth J. Northcott. Chicago: University of Chicago Press.
Oberhuber, Karl. 1 982. "Zum 'Passivum' im Sumerischen". in: W. Meid (ed), Sprach-
 Bibliography 297

wissenschaft in Innsbruck. Innsbruck. 129- 134.

Oppenheim, A. Leo. 197 1 . (letter to editor). CurrentAnthropology 12: 2 19-220.
Pettinato, Giovanni. 198 1 . "La pronuncia sumerica dei numeri da 1 a 10 in un testo lessi
cale di Ebla". AIUON41 : 141-143.
Poebel, Arno. 1923. Grunzuge der sumerischen Grammatik. Rostock: Selbstverlag des
Verfassers.
----- 1932. Das appo sitionell bestimmte Pronomen der 1 . pers. sing. in den westsemiti
schen Inschriften und im Alten Testament. Chicago: University of Chicago Press. AS
3.
----- 1935. "The Sumerian Genitive Element". AlSL 5 1 : 145-176.
Postgate, J.A. 1974. "Two Points of Grammar in Gudea". lCS 26: 1 6-54.
Powell, Marvin A. 1978. "Ukubi to Mother ... The Situation is Desperate". ZA 68:
1 63-195.
----- 198 1 . "Three Problems in the History of Cuneiform Writing: Origins, Direction of
Script, Literacy". VisibleLanguage 15: 4 19-440.
Renger, Johannes. 1967f. "Untersuchungen zum Priestertum in der altbabylonischen
Zeit". ZA 58 ( 1 967) 1 10- 1 88; ZA 59 ( 1969) 104-230.
Roberts, Jack. 1972. The Earliest Semitic Pantheon. Baltimore: Johns Hopkins Univer
sity Press.
Romer, W.H.Ph. 1965. Sumerische "Konigshymnen" der Isin-Zeit. Leiden: E.J. Brill.
----- 1973. "Sumerian". in: J. Hospers, A Basic Bibliogrrulyli of the Semitic Languages.
Leiden: E.J. Brill. 38-45.
4
----- 1983. Einfuhrung in die Sumerologk . Nijmegen: Katholieke Universiteit Nijme
gen.
Rosengarten, Yvonne. 1967. Repertoire commente des Signes Presargoniques Sumeriens
de Lagas. Paris.
Salonen, Armas. 1961. Die Turen des alten Mesopotamien. Helsinki: Annales academiae
scientiarum fennicae.
Schmandt-Besserat, Denise. 1982. "How Writing Came About". Zeitschrift far Papy
rologie und Epigraphie 47: 1 -5.
----- 1986. "The Origins of Writing: An Archaeologist's Perspective". Written Commu
nication 3: 3 1 -45.
Seux, M.-J. 1 967. Epithetes royales akkadiennes et sumeriennes. Paris: Letouzey et Ane.
Sjoberg, Ake and E. Bergmann. 1969. The Collection of the Sumerian Temple Hymns.
Locust Valley: J.J. Augustin. TCS 3.
Sjoberg, Ake et al. 1984f. The Sumerian Diction1ID'. Philadelphia: The Babylonian
Section of the University Museum.
Snell, Daniel. 1 982. Ledgers and Prices: Early Mesopotamian Merchant Accounts. New
Haven: Yale University Press. Yale Near Eastern Researches 8.
Sollberger, Edmond. 1952. Le systeme verbal dans les inscriptions "royales" pre-sargo
niques de Lagas. Geneva: Librairie E. Droz.
----- 1966. Business and Administrative Correspondence under the Kings of Ur. Locust
Valley: J.J. Augustin. TCS 1 .
298 Appendix 4

----- 1972. "Ur III Society: Some Unanswered Questions". CRAI 18: 185-189.
----- 1973. "Problems of Translation". Or 42: 158-161.
Steiner, Gerd. 1978. "Der Gegensatz 'eigenes Land' : 'Ausland, Fremdland, Feindland' in
den Vorstellungen des alten Orients". CRAI25: 633-664.
----- 1980. "The Vocalization of the Sumerian Verbal Morpheme /=ed/ and its Signi
ficance". JNES 40: 21-41.
----- 198 1 . "ljamty. und maru als verbale Kategorien im Sumerischen und im Akka
disc hen". RA 75: 1-14.
Steinkeller, Piotr. 1977. "Seal Practice in the Ur III Period". in: Gibson and Biggs
(1977) 41-53.
1979. "Notes on Sumerian Plural Verbs". Or48: 54-67.
1984. "The Sumerian Verb lugx (LUL)". SELl : 5-17.
1987a. (review of Pennsylvania Sumerian DictionID). lNES 46: 55-59.
1987b. "The Administrative and Economic Organization of the Ur III State: The
Core and the Periphery". in: Gibson and Biggs (1987) 19-41.
----- 1988. "On the Identity of the Toponym LV.SU(.A)". lAOS 108: 197-202.
Stola, R. 198 1-82. "Zu den sumerischen Entsprechungen des akkadischen Imperativs in
spaten zweisprachigen Texten". Af0 28: 79-91 .
Strommenger, Eva. 1980. "The Chronological Division of the Archaic Levels of Uruk
Eanna VI to IIIIIl: Past and Present". AlA 84: 479-487.
Sullivan, Brian. 1979. Sumerian and Akkadian Sentence Structure in Old Babylonian
LiterID Bilingual Texts. Cincinnati: Hebrew Union College [dissertation].
Thomsen, Marie-Louise. 1984. The Sumerian Language: An Introduction to its History
and Grammatical Structure. Copenhagen: Akademisk Forlag. Mesopotamia: Copen
hagen Studies in Assyriology 10.
Tigay, Jeffrey H. 1982. The Evolution of the Gilgamesh Epic. Philadelphia: The Univer
sity of Pennsylvania Press.
Vaiman, A.A. 1974. "Uber die Protosumerische Schrift". Acta Antiqua 22: 15-27.
van Driehl, G. 1982. "Tablets from Jebel Aruda". in: G. van Driehl (ed), Zikir umim:
Assyriological Studies Presented to F.R. Kraus. 12-25.
Vanstiphout, Herman L.J. 1985. "On the Verbal Prefix lif in Standard Sumerian". RA
79: 1-15.
Veenhof, K.R. 1986. "Cuneiform Archives. An Introduction". in: K.R. Veenhof (ed),
Cuneiform Archives and Libraries. Leiden: Nederlands Instituut voor het Nabije Oos
ten.
Walker, c.B.F. 1987. Cuneiform. Berkeley/London: University of California Press and
the British Museum.
Weadock, Penelope. 1975. "The giparu at Ur". Iraq 37: 101-128.
Westenholz, Aage. 1985. "An essay on the Sumerian 'Lexical' Texts of the Third Millen
nium". Or 54: 294-298.
Wiseman, D.J. 1960. "The Goddess Lama at Ur". Iraq 22: 166-171.
Woolley, Sir Leonard. 1982. Ur 'of the Chaldees'. Revised and updated by P.R.S.
Moorey. Ithaca: Cornell University Press.
 Bibliography 299

Yoshikawa, Mamoru. 1968a. "The Manl and IJamu Aspects in the Sumerian Verbal
System". Or 37: 401 -416.
----- 1968b. "On the Grammatical Function of -E- of the Sumerian Verbal Suffix -E-DEI-
E-DA(M)". JNES 27: 25 1-261.
1974. "The Maru-Conjugation in the Sumerian Verbal System". Or 43: 17-39.
1979. "Verbal Reduplication in Sumerian". ASJ 1 : 99- 1 19.
1980. "On the Alleged 'Ueberhlingende Vokale''' . ASJ2: 1 88- 195.
198 1a. "The Sumerian Verbal Prefix al-". CRAI28: 66-7 1 .
198 1b. "Plural Expressions in Sumerian Verbs". ASJ3: 1 1 1- 124.
Yushu, Gong. 1987. "ErgativiUit und das Sumerisch". Journal of Ancient Civilizations
2: 85- 120.
Zakar, Andras. 197 1 . "Sumerian-Ural-Altaic Affinities". Current Anthropology 12:
2 15-225.
Zettler, Richard. 1977. "The Sargonic Royal Seal: a Consideration of Sealing in Meso
potamia". in: Gibson and Biggs (1977) 33-39.
Zhi, Yang. 1987. "The Name of the City Adab". Journal of Ancient Civilizations 2:
121- 125.
Concordance of texts used
This concordance lists all the texts used in the Lessons. For each text, the following
information is given: number according to Hallo's catalogue (if available); publication of
cuneiform; publication of transliteration and translation in SAKI; publication of translation
in IRSA; location of photograph. The following abbreviations are used:
Cuneiform
BE = Babylonian Expedition of the University of Pennsylvania, Series A, Cuneiform
Texts. Philadelphia.
BE 1 1 896 (Hilprecht)
=

BIN = Babylonian Inscriptions in the Collection of James B. Nies. New Haven.

BIN 2 1920
=

CT = Cuneiform Texts from the British Museum. London.

CT 5 = 1 898
CT 2 1= 1905
CT 36 192 1
=

MDP = Memoires de la Delegation en Perse. Paris.

MDP 6 1905
=

OIP = Oriental Institute Publications. Chicago.

OIP 14 1930
=

PBS = Publications of the Babylonian Section. The Museum, University of

Pennsylvania. Philadelphia.
PBS 1 3 1922
=

PBS 1 5 1926
=

R = Rawlinson, H. The Cuneiform Inscriptions of Western Asia. London.

300 Appendix 4

I R = 1 86 1
UET = Ur Excavations. Texts. London.
UET 1 = 1 928
UET 3 = 1 937
UET 8 = 1 965
UVB = Vorlaufiger Bericht iiber die von der Deutschen Forschungsgemeinschaft in Uruk
Warka unternommenen Ausgrabungen. Berlin.
UVB 10 = 1 939
UVB 1 2/ 1 3 = 1 956
VAS = Vorderasiatische Denkmaler. Berlin.
VAS 1 = 1 907
YOS = Yale Oriental Series. New Haven.
YOS 1 = 1 9 1 5

Secondary Literature

Hall = H.R. Hall, A Season's Work at Ur. London, 1 930.

Hallo = W.W. Hallo, "The Royal Inscriptions of Ur: A Typology". HUCA 33 ( 1 962) 1 -
43 .
HSA = L. King, History of Sumer and Akkad. London, 1 9 1 0.
IRSA = E. Sollberger and J.-R. Kupper, Inscriptions royales sumeriennes et akkadiennes.
Paris, 1 97 1 .
SAKI = F. Thureau-Dangin, Die sumerischen und akkadischen Konigsinschriften.
Leipzig, 1 907.

Several individuals and publishers have either supplied photographs, or provided

permission to reproduce copyrighted material. I would expressly like to thank the fol
lowing: The British Museum, for supplying photographs of the drawing of the ziggurat in
Lesson 1 ; of the drawing of the ziggurat in Lesson 9; of the pedestal in Lesson 1 5 ; of the
Old Akkadian seal in Lesson 2 1 ; of texts CT 2 1 , 2: 90004; CT 2 1 , 3: 900 1 5; CT 2 1 , 6:
89 1 26; UET 1, 55. The Pierpont Morgan Library, for the photographs in Lesson 7 of the
cuneiform tablet and the canephore figures. The University Museum of the University of
Pennsylvania for the photographs of the stela in Lesson 1 and the stamp seals in Lesson 2.
Jack Finegan, for the photograph of the ziggurat in Lesson 1 (reproduced courtesy of
Westview Press). Corn ell University Press, to reprint from Ur 'of the Chaldees': A
Revised and Updated Edition of Sir Leonard Woolley's Excavations a t Ur by P.R.S.
Moorey 1 982 The Estate of Sir Leonard Woolley and P.R.S. Moorey. The Pontifical
Biblical Institute, for several quotations from Sumerian Grammar in B abylonian Theory by
Jeremy Black 1 984. Any inadvertent omission of acknowledgments will be corrected in
the future.

Textual Concordance
Text 1 Ur-Nammu 8
UET 1 , 40
Drawing of ziggurat: Courtesy of British Museum. Reproduced in Woolley
 Bibliography 301

( 1 982) 148
Photograph of ziggurat: Courtesy of Jack Finegan. Reproduced in Finegan
( 1 979) 5 1 . Reproduced courtesy of Westview Press
Photograph of stela: Courtesy of University Museum, University of
Pennsylvania
Text 2 Ur-Nammu 7i
CT 2 1 , 3 : 900 1 5
SAKI 1 86d
Photograph: Courtesy of British Museum. Reproduced in RSA, facing page
280
Photographs of stamp seals: Courtesy of University Museum, University of
Pennsylvania
Text 3 Ur-Nammu 9i
3a CT 2 1 , 2: 90004
Photograph: Courtesy of British Museum. Reproduced in Rall 1 06
3b CT 2 1 , 2: 90009
SAKI 1 86b
IRSA IIIA l c
Text 3c Ur-Nammu 7i
CT 2 1 , 3: 90006
SAKI 1 86d
Text 4 Ur-Nammu 3 1
UET 1 , 34
Text 4a Ur-Nammu 1 1
CT 2 1 , 5 : 9000 1
SAKI 1 96e
Text S Ur-Nammu 23i
UET 1 , 46
Text 6 Ur-Nammu 5ii
UET 1 , 4 1 a
IRSA IIIA lj
Text 7a Ur-Nammu 3iii
Hallo, in: Ancient Mesopotamian Art and Selected Texts. New York: The
Pierpont Morgan Library ( 1 976), p. 22
Photograph : Courtesy of The Pierpont Morgan Library. Reproduced in ibid, p.
17
Photograph of canephore figurines: Courtesy of The Pierpont Morgan Library.
Reproduced in ibid, p. 22
Text 7b Ur-Nammu 3ii
CT 2 1 , 4: 90802
SAKI 1 86g
Text 7c Ur-Nammu Si
CT 2 1 , 5 : 90296
302 Appendix 4

SAKI 1 86f
IRSA IIIAlj
Text 8 Ur-Nammu 1 6
BE 1 , 1 2 1
SAKI 1 88k
Text 8a Ur-Nammu 3i
BE 1, 1 22
SAKI 1 87, note h
Text 9 Ur-Nammu l Oi
CT 2 1 , 7: 90000
SAKI 1 86c
IRSA IIIAld
Drawing of ziggurat: Courtesy of British Museum; reproduced in Woolley
(1982) 234
Text 10 Shulgi 52
VET 1 , 55
Photograph of 30-mina weight: Courtesy of British Museum
Text lOa Ur-Nammu lOi
IR I i5
SAKI 1 86c
IRSA IIIAld
Text 1 1 Shulgi 29
CT 5, 2: 122 1 8
SAKI 194x
IRSA IIIA2u
Photograph: Hallo, in Denise Schmandt-Besserat (ed), The Legacy of Sumer.
Malibu: Undena Publications (1976), p. 133
Text l la Ur-Nammu Iv
Boson, Aegyptus 1 5 (1935) 420
IRSA IIIAl b
Text 12 (Ur-Nammu)
UET 8, 4:21
IRSA IIIAlk
Text 12a Shulgi 20i
PBS 15, 42
IRSA IIIA2g
Photograph is of Shulgi 22ii: McCown, Archaeology 5 (1952) 74
Text 1 3a Amar-Sin 2ii
CT 2 1 ,'24: 90034
SAKI 196b
IRSA IIIA3b
Text 1 3b Amar-Sin 2iii
VAS 1 , 26
 Bibliography 303

IRSA llIA3b
Text l 3e Amar-Sin 1
BE 1 , 22
SAKI 196a
IRSA IIIA3a
Text 14 Amar-Sin 5iv
Sumer 3 (1947) le, facing p. 236
IRSA llIA3h
Text 14a Vr-Nammu 22i
VET 1 , 45
Text 1 5 Amar-Sin 3i
er 2 1 , 25f: 908 1 1
SAKI 198d
IRSA IIIA3e
Photograph of pedestal: Courtesy of British Museum. VET 1 , 172, pI. V
Text 1 6 Amar-Sin 1 0
BE 1, 21
SAKI 1989
IRSA IIIA3f
Text 16a Shulgi 46
RO 2 (1925) 1 89
IRSA IIIA2f
Text 17 Amar-Sin 1 1
VVB 1 0, pI. 28
IRSA IIIA3d
Photograph: VVB 10, pI. 23a
Text 18: Shu-Sin 6i
CT 2 1 , 28: 90844
SAKI 200b
IRSA IIIA4e
Text 1 8a Shulgi 43
MDP 6, 22
SAKI 194y
Text 19 Shu-Sin 9iii
YOS 1 , 20
IRSA IIIA4d
Text 19a Amar-Sin 17
BIN 2, 17
Text 20 Shu-Sin 3
OIP 14, 43
IRSA IIIC1 b
Text 21a Shu-Sin 17
de Sarzee, Deeouvertes en Chaldee (1912) voI. 2 pI. 26-bis: 5
304 Appendix 4

SAKI 202e
IRSA IIIA4i
Text 21b Ibbi-Sin 7i
UET 3, 52
Photograph of Old Akkadian seal: Courtesy of British Museum.
Text 21c Ibbi-Sin 8iii
PBS 1 3, 5: CBS 12570
Text 21d: Shulgi 5 l iii
UET 1 , 287
Photograph: The Illustrated Bible DictionID.)' (Tyndale, 1980), vo!. 3, 1 635
Text 22 Shulgi 47
CT 2 1 , 9: 89 1 3 1
SAKI 194z
IRSA IIIA2w
Collation: Gadd, Iraq 1 0 (1948) 98 n. 1
Photograph: Courtesy of British Museum. Reproduced in RSA, facing p. 246
Text 22a Ur-Nammu 36
CT 2 1 , 6: 891 26
SAKI 1 88n
IRSA IIIA1m
Photograph: Courtesy of British Museum. Reproduced in RSA, opposite p.
246
 Appendix 5
Further Work

The Ur III royal inscriptions studied in this book present only a limited picture of
S umerian. For example, not all the modal-prefixes occur, and there is only limited use of
the first and second persons. In order to deepen one's knowledge of S umerian, it is neces
sary to do two things: read well edited Sumerian texts, and read the most important and/or
recent secondary literature about Sumerian.
More "variety" in grammar occurs in Sumerian literary texts, and for that reason they
are perhaps the most useful texts to read at this stage. However, this is not as easy as it
sounds. Because most literary texts were copied down in the Old Babylonian period or
later, they are often influenced by Akkadian grammar, and may contain forms which are
simply "wrong" by the normative rules of Classical Sumerian grammar; such wrong forms
can be disconcerting to a relati ve beginner.
To obviate this problem, it is necessary to work through literary texts which are well
edited. However, some Sumerologists are less interested in grammatical matters than other
S umerologists, and so may not discuss such matters as, for example, a seemingly incorrect
use of a personal-affix. Some scholars are more interested in lexicographical matters, and
may devote seemingly inordinate amount of space for citing all references for particular
words, instead of focusing on the grammar (given the lack of a complete up-to-date
dictionary of S umerian, this is often necessary).
In general, it is always valuable to first skim through an edition of a literary text, in
order to determine what kind of emphasis the modern-day editor is placing in his
commentary - is it primarily a grammatical commentary, or lexicographic, or stylistic, etc.
When reading through a new text, one will encounter variations of constructions seen
previously, or completely new constructions. These may or may not be discussed in a
commentary. Upon encountering a new form or construction, the first step is to isolate the
problem, that is, determine where in the grammar the problem lies: is it a previously unseen
modal-prefix, or a strange use of a dimensional-prefix, or a problem in a temporal clause,
etc. At this stage, one must turn to the standard grammars, and to other secondary
li terature.
The most general grammars which can be examined are, in chronological order (these
are further discussed below): Poebel ( 1 923); Falkenstein ( 1 959); Falkenstein ( 1 978 2 );
Romer ( 1983 4 ); Thomsen ( 1 984). Thomsen will be of most value to the student; she also
includes references to other secondary literature.
In addition to these more general grammars, there are certain articles, on specific
aspects of the grammar, which are always worth consulting. These are referred to in
Tomsen, but it is useful to keep photocopies of them at hand. Full references are given in
theBibliography.

moods: Jacobsen ( 1 965); Edzard ( 1 9 7 1 f); Michalowski ( 1 980a); Kienast ( 1 9 8 1 b)

conj ugation-prefixes: J acobsen ( 1965); Gragg ( 1968; 1 972a)

305
306 Appendix 5

dimensional-prefixes: Gragg (1973a); Yoshikawa (198 1a)

personal-affixes: Michalowski (1980a)
ijamty manl:
- Yoshikawa ( 1968a; 1968b; 1974); Edzard (1971f); Michalowski
(1980a); Steiner (1981)
plural verbs: Krecher (1967-8); Steinkeller (1979); Yoshikawa (198 1 b)
imperatives: Michalowski ( 1980a)
copula: Gragg (1968)
subordinate clauses: Gragg (1973b)
relative clauses: Gragg (1972)
ergativity: Foxvog (1975); Michalowski ( 1980a); van Aalderen (1982)
As stated above, in addition to reading more texts, it is necessary to read more about
the language itself. At this point, the student can profit by a systematic reading of
Thomsen. Thomsen cannot practically be used without some previous knowledge of
Sumerian, but this Manual should provide enough background to make Thomsen useable.
The articles listed above, besides being consulted for reference when an unfamiliar form is
encountered, should also be read in their entirety.
For further bibliography on grammatical subjects, one can consult: Sollberger (1952),
with a rather complete bibliography through 1952; R6mer (1973); Falkenstein (19782 );
R6mer (19834 ); and Thomsen ( 1984).
There have been five modem larger-scale grammars of Sumerian. The first of these
was Poebel (1923). Poebel was perhaps the first person to really understand Sumerian
grammar, and to put its study on a sound footing. Although written in a very old-fashioned
style, and although wrong in a number of details, it is still a valuable tool.
Falkenstein published a complete "sketch" of Sumerian in 1959. This work is often
cited by Sumerologists, but it suffers from being too terse (the section on morphology and
syntax occupies less than thirty pages), and from a rather whole-sale use of concepts and
terms which are more fitting for the description of Indo-European or Semitic languages. It
also has a rather confusing mixture of synchronic and diachronic description. However,
since Falkenstein's views differ somewhat from those presented in this book, and since his
views are accepted (sometimes only tacitly) by many Sumerologists, this book should be
worked through.
Falkenstein has also produced a detailed grammar of the language of the inscriptions
of Gudea (second edition, prepared posthumously, 1978). Although limited to one group
of texts, this is a very useful work. There are three volumes: (1) script and morphology;
(2) syntax; and (3) an introduction to the texts, discussing historical background, iden
tifying divine names, etc. This introduction is handy for identifying proper names oc
curring in other texts. The morphology volume contains elaborate paradigms; these are
quite useful, and are in general more accurate than those in Poebel.
 Further Work 307

Romer ( 19834 ) is essentially an updated Falkenstein; the section entitled "Einiges zur
Sprache" is less than fifty pages long. However, it has an enormous amount of
bibliographic references - there are 641 footnotes; these references are what make the work
valuable. A new edition is being prepared.
Finally, the most recent grammar is that of Thomsen ( 1984); there is an important
review by Jacobsen (1988). Thomsen bases her work on the Old Babylonian literary texts,
because these exhibit the greatest variety of grammatical variation. The danger with this
approach is the fact that in these texts one must always be on guard against Akkadian
influence. In spite of this methodological criticism, Thomsen contains a balanced,
thorough, and relatively uncontroversial presentation of the grammar, more-or-Iess in the
Falkensteinian tradition. Her book would be difficult for anyone who does not know the
principles of cuneiform script, but for those who already know some Sumerian, it will
serve as a standard reference for some time.
This Manual will be followed by a second volume consisting of extracts from Inanna's
Descent, with a commentary focussing on the grammar of the text.
 Appendix 6
Topical Index

The index is designed to tie together information scattered throughout this book, in
order to permit quick reference and review. Only significant discussions are listed. Re
ferences are to pages.

ablative case 177, 197

absolute case 34
abstract nouns 67, 105, 25 1
adjectives 50, 82, 84, 97, 208, 209, 217
adnominal cases 54
adverbial cases 54, 68, 138
Amar-Sin 154, 209
amissability 3 1 , 83, 121
animacy 32, 39, 1 34, 173
anticipatory genitive 1 1 1 , 170- 1 7 1 , 25 1
article 3 1
aspect 37, 133- 1 34
Auslaut 3 1 , 32, 175
autograph 30, 234
benefactive 32, 1 1 1 , 200, 230, 25 1
brick 41, 91
brick-stamp 52-53
building inscription 4 1 , 62, 77, 78, 9 1 , 230
case system 32, 53, 54
circumstantial clause 1 98, 199
cohortative 175, 252
comitative case 1 35, 17 1- 172
compound logogram 74, 77
compound nouns 101, 105
compound verbs 67, 67-68, 70, 95, 1 29- 1 30, 1 36, 138, 176-177
concrete nouns 69, 105, 25 1
cone 77
conjugation 40, 139, 202
conjugation-prefix 34, 39-40, 88-89, 90, 136, 1 37-1 38, 1 38, 175, 200, 202-203, 252
ba 136, 1 37, 219, 220
bi 133, 1 37, 138, 219
1 34, 88, 90, 175, 200
im 200, 202
mu 34, 35, 88
coordination of nouns 5 1 , 1 85, 1 87- 1 88

309
3 10 Appendix 6

coordination of verbs 6 1 , 62, 1 88

copula 1 22, 170, 174
cuneiform sign formation 26-27 , 6 1 -62, 73-74, 89, 1 04, 1 1 8 , 1 25 , 1 53 , 1 74- 175, 1 82,
1 9 3 , 1 95, 2 1 2-2 1 3, 2 1 3, 223, 224, 247-248, 252
curse-fonnula 1 33, 1 42- 1 43 , 1 7 3 , 1 90
dative case 32, 1 60- 1 6 1 , 25 1
demonstrative 1 35
desiderative 1 36, 1 7 3 , 252
dimensional-prefix 35, 68, 1 76, 200
ablative 177, 1 97
comitative 1 36, 1 38, 170- 1 7 1 , 200
dative 35, 200
locative 82-83 , 84, 1 99, 200
locative-tenninative 1 04, 2 1 8
terminative 68, 69, 70
divinity of Ur III kings 1 1 1 , 1 1 3
door socket 95, 97-98, 1 45, 1 88
double genitive 1 52, 2 1 6
Ebla 77, 1 1 1 , 1 27, 1 88
.ed 1 85- 1 86, 1 87
enclitic copula 1 22, 1 70, 1 77, 198
ergative case 33, 1 04, 1 62, 1 99, 203 , 227
ergativity 33, 1 40- 1 4 1 , 202f
foundation deposit 9 1
function of texts 4 1 , 9 1 , 97-98, 1 8 8 , 254
gender 27, 3 1
genitive 3 1 , 33, 39, 5 1 -52, 54, 54-55, 76, 1 1 1 , 1 2 1 , 123, 1 52, 229
gunu 1 04, 193, 252
b amt!! 37, 40, 1 33
lbbi-Sin 23 1 , 242
indicative 34
intransitive 197, 1 9 8 , 202
loan words 74, 77, 80, 82, 83, 1 03 , 1 06, 1 24, 1 27 , 1 58 , 175, 2 1 2, 2 1 3, 2 1 9 , 228, 247
locative case 82, 84, 1 35
locative-terminative case 1 03- 1 04, 2 1 8, 2 1 9 , 235, 25 1 , 25 1 -252
maru 37 , 1 3 3- 1 34, 1 39f, 1 7 1 , 172
maru-suffix 1 34, 1 39f
Mesannepadda construction 1 54
modal-prefix 34, 1 36, 1 37 , 1 40, 1 7 5 , 1 85, 252
negation 1 85 , 198
nominal phrase 32, 35, 1 60
nominalizer 50, 82, 8 8 , 97, 208
numbers 1 09, 23 1 , 232, 233
 Index 311

optional pronominal-prefix 1 7 1 , 1 72, 176

overhanging vowel 95, 1 23- 1 24, 1 26, 1 52, 173
palaeography 49, 52, 60, 89, 1 1 6, 1 49, 227, 234, 256
participle 50, 82, 1 22, 1 54, 1 87, 208
passive 1 05 , 1 97- 1 98, 202
personal-affix, pre-verbal root 35, 89, 1 34, 1 4 1 , 1 72, 173, 1 86, 1 87, 2 1 8, 252
personal-affix, post-verbal root 36, 1 34, 1 36, 1 4 1 , 252
plural - nouns 76, 82, 2 1 7
plural - verbs 1 72
possessive-suffix 32, 97, 103, 235-236
proper name fonnation 27, 33, 1 22, 1 22- 1 23, 1 70, 1 77, 1 86, 2 1 7, 236, 246
reduced relative clause 1 5 1 - 1 52, 1 54
reduplication 1 39, 1 4 1
relative clause 88, 1 32- 1 33, 1 5 1 , 1 85, 1 86, 208, 209
restoration of text 207, 234, 246, 250
root 36, 1 36, 198, 200, 252
seal inscriptions 235, 236f, 240
SeSSig 252
Shulgi 1 1 4- 1 1 5, 1 54, 232
Shu-Sin 209, 220, 230, 242
standard inscription 9 1 , 1 1 2, 1 52
subordinate clause 1 32
substrate 1 06, 1 5 3 , 1 5 8
syllabic structure 3 3 , 1 2 1 -2
syllabic writings 252-253
temporal clause 1 32- 1 33, 2 1 7
tense 37, 1 33- 1 34, 1 85
terminative case 67, 69, 1 03, 2 1 7
titulary 44-45, 56, 1 1 3- 1 1 4, 1 6 1 , 1 63, 1 97, 208
!! 1 85
Ur-Nammu 44, 56, 1 1 3 , 1 1 4
verbal phrase 34, 37, 3 8
vocalic assimilation 7 6 , 1 24- 1 25, 223 , 228
vocative 235, 25 1
votive inscription 70-7 1 , 1 42, 254
votive objects 70, 1 1 7, 1 42- 1 43 , 1 77, 2 1 0, 22 1 , 240, 254
weights 1 1 2, 23 1 , 232, 232-233 , 243
word order 38, 52, 1 6 1 , 2 1 7, 2 1 8, 230, 24 1
word order typology 40-4 1 , 55, 1 1 1 - 1 1 2
year-dates 2 1.9-220, 22 1
Pllhlirafi()ns of lI.1fAS
nu' Infefllationol lmtituf(' for .\fl'soporamiol1 Irea Sflldics

PUBLICATIONS OF THE TERQA ARCHAEOLOGICAL PROJECT

Modular Preli
minar)' Reports

TPR I G. BUCCELLATI and M. KELLYBUCCELLATI

General Introduction and the Strarigraphic Record of the Firsl Two Se,ISOIlS
llll Sfudies 1/3 (1977).
Syru-lIfel"opowmi

TPR 1 G. BUCCELLATI
A Cuneiform Tablet from !he Second Season. Syro.lft'.wpowmwn Swdies 1/4 (1977).
TPR J M. KELLY-BUCCELLATI and L. MOUNTWILUAMS
Object Typology of Ihe Second Season (Excluding Vessels ,lnd Lithics)
SyrrrMewpatomiall Sflldif'. 1/5 (1917).

TPR .j. M. KELLY-BUCCELLATI and W. R. SHELBY

Ol'fOmic Ves>el TYpology of the First and Second Seosolls, S.)'ro-Meoporomal/
f SllIf/ies 1/6 ( 1 977).

Tl'R s: A. MAHMOUD
Die Induslrie der islamischen KeraOlik ous der lweilen Season
SyrfJ-MeWl/lofOmioll Sfudi,',\ 2/5 (1 978).

TPR 6 G. BUCCELLATI and M. KELLY-BVCCELLATI

Chronicle and Srrnlillraphk Record of Ihe Third Season. Syro.llesopotolllklll Swdill 2/6 t l 978).

TPR 7 O. ROUAULT
Les Documents epigraphiques de la Iroisieme SIIison, Syro-Mf''opotollliall SmdieI 2/7 ( 1979),

Tl'R 8 L. MOUNTWILLiAMS
Objecl Typology of the Third Season: The Third and Second Millennia
S.l'rl).M"SIIfJf!lomiall Sf/uJies 3/2 (1 980).

TPR 9 E, GRIFFIN and W. R. SIiELBY

Cramic Veso;el Typology of the Third Season, s.1ru-Jlf>sIJpmomialJ Swdi<'s (forthcoming).

IPR 10 G. BUCCELLATI I,ilh Ihe collnbornlion of M. KELLY-BUCCELLATI and J, KNUDSTAD

The Fourth Sea_oil: Introduction and the Slrnlilll"aphic Record
Bibli/llhera Mesol}(ltomira Vot. 10 t l 979).

TPR 1 1 N . M. MAGA LOUS IS, N al.

Soureing Terhllique for Ceramics and Soils. SI"TO-.lleso{lotomialJ Srudie.1 3/'" ( 1 91\0),

IPR 1 2 G. BUCCElLATI Hlld 0, ROUAULT

Digital PJolling of Archllt."ological Floor Pllln
k(/ ReSf'{lrcll in Nl'Or Eu.'ft'fII Stlllllt>s) 1 / 1 t l 983).
CAR.WS (CulII/llllrr .4it

FinflJ Reports

TFK I o ROlIAl!IT, I'Arthi>e dt." PUlurum. Bib/it/fflua Mewpt/lamiro Vol. 16 ( 1 9"4).

R I \ 'II I"S \IJ)S ,\NB RISb\RCH TOOLS IN ANClf:NT NEAR I\SUR;\; STlIOll:S

U SIU.NNUM. 1llati/"'l}pti.n Inde\ of "'ouu."rr ('<lIICUt> Diclimrary of "'-ddf" f/{I"plion (I 977)

J 11 SNf-ll.. A WorLbuulo. uf (uneifonn Sian ( 1 919)

.. J M SASSON. 'lIIled Tnh rrom MIITi: A TabuJalion ( (9NO)

IWI>I'.'NA PI'BLle4 rUJN,"'

P,Q, Do .. 97, Malibu, California, U.SX (}():!bS